Optical cable external breakage risk prediction method and system, electronic equipment and storage medium

Abstract

The invention relates to the field of optical cable data operation and maintenance, in particular to an optical cable external breakage risk prediction method and system, electronic equipment and a storage medium, and the method comprises the steps: constructing a risk prediction knowledge graph body according to optical cable external breakage risk knowledge; constructing a local risk prediction knowledge graph through historical external damage influence data of each optical cable region, and generating a prediction risk training corpus through the local risk prediction knowledge graph; performing fine tuning training on the pre-trained semantic model through the prediction risk training corpus; the method comprises the following steps: acquiring real-time external breakage influence data of an optical cable area, inputting the real-time external breakage influence data into a fine-tuned semantic model, predicting a candidate external breakage risk source of the optical cable area, and acquiring a target external breakage risk source according to candidate sound wave data and real-time sound wave data of the candidate external breakage risk source. Compared with the prior art, the method has the advantages that the pre-trained semantic model is finely adjusted by using the formed local risk prediction knowledge graph, so that the pre-trained semantic model can effectively realize prediction and identification of the optical cable external damage generation source.

Description

Technical Field

[0001] This invention relates to the field of optical cable data operation and maintenance, and more specifically, to a method, system, electronic device, and storage medium for predicting the risk of external damage to optical cables. Background Technology

[0002] With the continuous development of optical cable technology, the difficulty of optical cable operation and maintenance is constantly increasing. In order to reduce the occurrence of external damage events, it is usually necessary to predict the external damage events that may occur around the optical cable. In the existing technology, the identification of external damage events of optical cables is usually done by collecting vibration information or sound wave information received by the optical cable, and then analyzing the vibration information or sound wave information to identify external damage events around the optical cable. Existing analytical methods have certain shortcomings: Firstly, they typically rely on a single acoustic signal to identify external damage events, making comprehensive identification and prediction difficult. Furthermore, when the optical cable is in a complex environment, numerous factors can influence the judgment, leading to inaccurate predictions based solely on acoustic information. Secondly, existing technologies analyze the entire optical cable, but optical cables span vast areas with distinct regional characteristics, such as high traffic volume or frequent construction sites. Different regions pose varying risks of external damage, and existing technologies cannot predict and identify these regional characteristics. Thirdly, current technologies often utilize neural networks to analyze cable data and directly predict external damage events. This often only provides the prediction result, without understanding the prediction process or the underlying causes, hindering prevention. Summary of the Invention

[0003] The present invention aims to overcome at least one of the defects (deficiencies) of the prior art and provide a method, system, electronic device and storage medium for predicting the risk of external damage to optical cables, which can effectively improve the accuracy of identifying the source of external damage events.

[0004] To solve the above-mentioned technical problems, the first technical solution proposed by this invention is: a method for predicting the risk of external damage to optical cables, the method comprising the following: Construct a risk prediction knowledge graph ontology based on knowledge of optical cable external damage risks; Historical data on the impact of external damage to optical cables in various regions of the optical cable are obtained based on knowledge of external damage risks. These regions are defined according to the pre-defined impact range of each optical cable segment. Based on the risk prediction knowledge graph ontology, a local risk prediction knowledge graph corresponding to each optical cable area is constructed according to the historical external damage impact data of each optical cable area. Generate corresponding prediction risk training corpus based on the local risk prediction knowledge graph; The pre-trained semantic model is fine-tuned based on the prediction risk training corpus corresponding to each optical cable area to obtain the fine-tuned semantic model; the fine-tuning training specifically involves the pre-trained semantic model using the prediction risk training corpus to predict external damage risk sources for the corresponding optical cable area. Real-time external damage impact data and real-time acoustic data of the optical cable area where the external damage event occurred are obtained; the real-time acoustic data is extracted and restored based on the optical cable data of the corresponding optical cable area. The real-time external damage impact data is predicted using the fine-tuned semantic model to obtain several candidate external damage risk sources for the corresponding optical cable area; Based on the candidate external damage risk sources, obtain the corresponding candidate acoustic wave data, compare each candidate acoustic wave data with the corresponding real-time acoustic wave data, and obtain the target external damage risk sources for each optical cable area.

[0005] Furthermore, the construction of the risk prediction knowledge graph ontology based on the knowledge of optical cable external damage risk specifically includes: Knowledge extraction is performed on the knowledge of external damage risk of optical cable to obtain the entities and relationships of the knowledge of external damage risk of optical cable; Based on the entities and relationships, the structure of the risk prediction knowledge graph ontology is constructed to obtain the risk prediction knowledge graph ontology.

[0006] Furthermore, the historical external damage impact data includes several historical external damage risk sources, as well as historical risk source information corresponding to the historical external damage risk sources. The historical external damage risk sources are classified into several risk types according to the risk prediction knowledge graph ontology. The real-time external damage impact data includes several real-time external damage risk sources, as well as real-time risk source information corresponding to the real-time external damage risk sources. The real-time external damage risk sources obtain their corresponding risk types based on the risk prediction knowledge graph ontology.

[0007] Furthermore, the construction of a local risk prediction knowledge graph for each optical cable region based on the risk prediction knowledge graph ontology and historical external damage impact data of each region specifically includes: Based on the entities in the risk prediction knowledge graph ontology, the historical external damage risk sources corresponding to the historical risk types are obtained from the historical external damage impact data of each optical cable area; Generate specific instances of corresponding entities based on the corresponding historical external risk sources and historical risk source information; Based on the specific instances and corresponding entities, a local risk prediction knowledge graph for the corresponding optical cable area is constructed.

[0008] Furthermore, the step of acquiring real-time external damage impact data for each optical cable area and predicting the real-time external damage impact data using the fine-tuned semantic model specifically includes: Based on the real-time external damage impact data and the real-time risk source information, a specific instance of the corresponding entity in the corresponding local risk prediction knowledge graph is generated, and the corresponding local risk prediction knowledge graph is updated based on the generated specific instance. Based on specific instances of the updated local risk prediction knowledge graph, several real-time prediction semantic features are constructed. The fine-tuned semantic model is used to predict several real-time predictive semantic features of the corresponding optical cable area.

[0009] Furthermore, the step of generating corresponding prediction risk training corpus based on the local risk prediction knowledge graph specifically includes: Based on each specific instance in the local risk prediction knowledge graph, construct corresponding semantic features of external damage impact; Based on the relationships between entities corresponding to specific instances in the local risk prediction knowledge graph, the semantic features of external damage impact are integrated to obtain several sets of semantic features of external damage impact. Add corresponding external damage event tags to each of the aforementioned sets of semantic features related to external damage impact; Based on the semantic feature set of external damage impact after adding external damage event tags, construct the prediction risk training corpus corresponding to each local risk prediction knowledge graph, and add corresponding region tags to each prediction risk training corpus.

[0010] Furthermore, the step of fine-tuning the pre-trained semantic model based on the predicted risk training corpus corresponding to each optical cable region to obtain the fine-tuned semantic model specifically includes: The pre-trained semantic model is fine-tuned based on the predicted risk training corpus corresponding to each optical cable area, and several fine-tuned semantic models are obtained through the fine-tuning training; the number of fine-tuned semantic models matches the number of optical cable areas.

[0011] To solve the aforementioned first technical problem, the second technical solution proposed by this invention is: a fiber optic cable external damage risk prediction system equipped with the aforementioned first technical solution, the system comprising: The knowledge graph ontology construction module is used to construct a risk prediction knowledge graph ontology based on knowledge of optical cable external damage risks. The historical data acquisition module is used to acquire historical external damage impact data for each optical cable area on the optical cable based on the knowledge of external damage risk; the optical cable area is divided according to the influence range of each preset optical cable segment on the optical cable; The local knowledge graph construction module is used to construct a local risk prediction knowledge graph corresponding to each optical cable area based on the risk prediction knowledge graph ontology and the historical external damage impact data of each optical cable area. The training corpus generation module is used to generate corresponding prediction risk training corpus based on the local risk prediction knowledge graph. The model training module is used to fine-tune the pre-trained semantic model according to the prediction risk training corpus corresponding to each optical cable area, so as to obtain the fine-tuned semantic model; the fine-tuning training specifically refers to the pre-trained semantic model using the prediction risk training corpus to predict the external damage risk sources of the corresponding optical cable area. The real-time data acquisition module is used to acquire real-time external damage impact data and real-time acoustic wave data of the optical cable area where the external damage event occurred; the real-time acoustic wave data is extracted and restored based on the optical cable data of the corresponding optical cable area. The candidate risk source acquisition module is used to predict the real-time external damage impact data through the fine-tuned semantic model and acquire several candidate external damage risk sources for the corresponding optical cable area. The target risk source acquisition module is used to acquire corresponding candidate acoustic wave data based on the candidate external damage risk sources, compare each candidate acoustic wave data with the corresponding real-time acoustic wave data, and acquire the target external damage risk sources for each optical cable area.

[0012] To solve the above-mentioned technical problems, the third technical solution proposed by the present invention is: an electronic device, comprising: Memory, used to store one or more computer programs; The processor, when the one or more computer programs are executed by the processor, implements the optical cable external damage risk prediction method described in the first technical solution above.

[0013] To solve the above-mentioned technical problems, the fourth technical solution provided by the present invention is: a computer-readable storage medium storing computer instructions, which are used to cause a processor to execute the optical cable external damage risk prediction method described in the first technical solution.

[0014] The beneficial effects of this invention are as follows: This application provides a method, device, electronic device, and computer storage medium for predicting the risk of external damage to optical cables. By constructing a risk prediction knowledge graph ontology using knowledge of external damage to optical cables, and using the risk prediction knowledge graph ontology as a framework, corresponding local risk prediction knowledge graphs are constructed for each optical cable region based on historical external damage impact data. This allows the local risk prediction knowledge graphs to effectively reflect the regional characteristics of the corresponding optical cable region based on the knowledge of external damage to optical cables. Consequently, the pre-trained semantic model can learn the regional characteristics of each optical cable region, and thus can make targeted predictions of the external damage risk situation of each optical cable region. Meanwhile, the constructed local risk prediction knowledge graph can effectively mine the relationships between various external damage-related entities based on the knowledge of optical cable external damage risk, and use a pre-trained semantic model for learning and training, so that the pre-trained semantic model can predict external damage events based on external damage-related entities. At the same time, it can also use the relationships between entities to determine candidate external damage risk sources, and then accurately determine the target external damage risk source from the candidate external damage risk sources by comparing acoustic wave data.

[0015] In addition, by utilizing the principles of knowledge graphs and combining them with semantic models to analyze historical and real-time external damage impact data, the analysis process of the semantic model becomes visible to users, improving the interpretability of candidate external damage risk source predictions and facilitating error correction by staff. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the optical cable external damage risk prediction method provided in this embodiment.

[0018] Figure 2 This is a schematic diagram illustrating the process of constructing the risk prediction knowledge graph ontology provided in this embodiment.

[0019] Figure 3 This is a schematic diagram illustrating the process of constructing a local risk prediction knowledge graph provided in this embodiment.

[0020] Figure 4 This is a schematic diagram illustrating the process of generating the prediction risk training corpus provided in this embodiment.

[0021] Figure 5 This is a schematic diagram of the prediction process of the fine-tuned semantic model provided in this embodiment.

[0022] Figure 6 This is a schematic diagram of the functional modules of the optical cable external damage risk prediction system provided in this embodiment.

[0023] Figure 7 This is a schematic diagram of the structure of the electronic device provided in this embodiment. Detailed Implementation

[0024] The accompanying drawings are for illustrative purposes only and should not be construed as limiting the scope of this application. To better illustrate the following embodiments, some components in the drawings may be omitted, enlarged, or reduced, and do not represent the actual dimensions of the product; it is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0025] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0026] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0027] The specific embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0028] like Figure 1 As shown in the figure, this embodiment provides a method for predicting the risk of external damage to optical cables, which may include the following steps: S1: Construct a risk prediction knowledge graph ontology based on knowledge of optical cable external damage risk; In this embodiment, the knowledge of optical cable external damage risk may include, but is not limited to, historical accident reports of optical cables (inspection analysis, fault analysis reports), expert experience analysis reports (risk assessment rules, case analysis), industry standards and specifications (optical cable protection specifications), etc. In this embodiment, as Figure 2 As shown, the construction of the risk prediction knowledge graph ontology may include the following steps: S11: Extract knowledge from the external damage risk knowledge of the optical cable to obtain the entities and relationships of the external damage risk knowledge of the optical cable; S12: Based on the entities and relationships, construct the structure of the risk prediction knowledge graph ontology to obtain the risk prediction knowledge graph ontology.

[0029] By utilizing the principles of knowledge graphs to extract knowledge about the risks of external damage to optical cables, we can effectively uncover various potential influencing factors that may lead to such incidents.

[0030] For example, the obtained entities can be represented as: optical cable segment (region, length, fiber core type), risk source (fixed source: building, bridge; mobile source: vehicle, aircraft; environmental source: weather, geology), etc.; the relationship between the obtained entities can be represented as: (optical cable segment) [distance] (risk source), (risk source) [type] (risk type), (environmental source) [impact] (optical cable segment), etc.

[0031] S2: Obtain historical data on the impact of external damage to each area of ​​the optical cable based on knowledge of external damage risks; In this embodiment, the optical cable area is divided according to the influence range of each preset optical cable segment on the optical cable; preferably, the environmental influencing factors in the corresponding optical cable areas should be the same or similar, such as one optical cable area being all urban areas and another optical cable area being all mountainous areas. The same or similar environmental influencing factors in the optical cable areas can ensure that the regional characteristics of the corresponding optical cable areas remain consistent to a certain extent, which facilitates the analysis of the corresponding optical cable areas.

[0032] In this embodiment, the historical external damage impact data includes several historical external damage risk sources and historical risk source information corresponding to the historical external damage risk sources. The historical external damage risk sources are divided into several risk types according to the risk prediction knowledge graph ontology. In this embodiment, the historical external damage risk sources can be understood as specific risk objects that could cause optical cable damage events. The specific risk objects can cover a very wide range, including major aspects such as road renovation, heavy rain, high temperatures, and peak traffic, and minor aspects such as electric picks striking the cables. It is understood that the principles of knowledge graphs can be used to mine the relationships between entities layer by layer, thereby achieving coverage of various factors and thus more comprehensively obtaining the historical external damage risk sources that will affect optical cable damage. The historical risk source information consists of detailed data parameters of the corresponding historical external damage risk source. Taking heavy rain as an example, the historical risk source information may include the duration of rainfall, rainfall amount, wind force, etc.

[0033] Understandably, the risk prediction knowledge graph ontology is constructed in step S1. The risk prediction knowledge graph ontology contains entities constructed based on various knowledge related to external damage to optical cables. Therefore, the historical external damage risk sources can be divided according to the entities in the risk prediction knowledge graph ontology to obtain several historical risk types, which facilitates subsequent corresponding queries for specific historical external damage risk sources.

[0034] S3: Based on the risk prediction knowledge graph ontology, construct a local risk prediction knowledge graph corresponding to each optical cable area according to the historical external damage impact data of each optical cable area; In this embodiment, the risk prediction knowledge graph ontology constructed in step S1 can be understood as the framework structure of the knowledge graph, which does not contain specific instances. The historical external risk sources obtained in step S2 are specific risk objects, which can be used as specific instances of the knowledge graph and together with the risk prediction knowledge graph ontology, they form a complete knowledge graph.

[0035] Specifically, such as Figure 3 As shown, the construction of the local risk prediction knowledge graph may include the following steps: S31: Based on the entities in the risk prediction knowledge graph ontology, obtain the historical external damage risk sources of the corresponding historical risk types from the historical external damage impact data of each optical cable area; As described above, the historical external damage impact data is divided into several risk types based on the entities in the risk knowledge graph ontology. Therefore, the corresponding historical risk type can be obtained through the entities in the risk prediction knowledge graph ontology.

[0036] S32: Generate specific instances of corresponding entities based on the corresponding historical external risk sources and historical risk source information; Specifically, in this step, the historical external risk sources can be used as objects of specific instances of the knowledge graph, and the corresponding historical risk source information can be used as attributes of the specific instances to generate specific instances of the knowledge graph.

[0037] S33: Based on the specific instance and the corresponding entity, construct a local risk prediction knowledge graph for the corresponding optical cable area.

[0038] Specifically, after obtaining the specific instance and attributes corresponding to the entity, the attributes of the relationship between entities can be obtained to improve the knowledge graph.

[0039] As described above, in this embodiment, the historical external damage impact data corresponds to a fiber optic cable region. Therefore, using the risk prediction knowledge graph ontology as a framework, the local risk prediction knowledge graph constructed using the historical external damage impact data of the corresponding fiber optic cable region can reflect various potential influencing factors of external damage to the fiber optic cable in the corresponding fiber optic cable region.

[0040] S4: Generate corresponding prediction risk training corpus based on the local risk prediction knowledge graph; In this embodiment, as Figure 4 As shown, the generation of the prediction risk training corpus may specifically include the following steps: S41: Construct corresponding semantic features of external damage impact based on each specific instance in the local risk prediction knowledge graph; In this embodiment, a corresponding text structure can be generated according to a preset format based on the instance object and attributes of the specific instance, and the text structure can be used as the semantic feature of the external impact of the specific instance.

[0041] S42: Based on the relationships between entities corresponding to specific instances in the local risk prediction knowledge graph, integrate the semantic features of external damage impact to obtain a set of semantic features of external damage impact; Since the relationships between entities in the local risk prediction knowledge graph are based on the knowledge of optical cable external damage risk, the relationships between entities can reflect the potential risks of corresponding external damage events. For example, if road renovation work that is too close to the optical cable may affect the optical cable, then the semantic features of the external damage impact corresponding to the road renovation and the semantic features corresponding to the optical cable can be integrated according to the relationship and relationship attribute of "distance not exceeding a preset distance value" to obtain the set of semantic features of the external damage impact. Each set of semantic features of the external damage impact can reflect a risk factor of an external damage event.

[0042] Understandably, based on the above, we can further integrate the specific instances corresponding to entities at multiple levels by exploring the relationships between entities. For example, in the road resurfacing example above, we can obtain the specific tools used for road resurfacing, such as pavers, hammers, and small bulldozers, as semantic features of external damage impact. Similarly, we can obtain all the relevant specific instances, obtain the corresponding semantic features of external damage impact, and integrate them to obtain a set of semantic features of external damage impact.

[0043] S43: Add corresponding external damage event tags to each of the aforementioned sets of semantic features related to external damage impact; Specifically, the external damage event tag can be a text description of the actual occurrence of an external damage event, such as fiber optic cable breakage or fiber optic cable sheath corrosion.

[0044] As described above, each set of semantic features related to external damage can reflect the risk factors of an external damage event. By adding external damage event labels to the corresponding sets of semantic features related to external damage, the model can be trained based on the corresponding external damage event labels.

[0045] S44: Based on the semantic feature set of external damage impact after adding external damage event tags, construct the prediction risk training corpus corresponding to each local risk prediction knowledge graph, and add corresponding region tags to each prediction risk training corpus.

[0046] In this embodiment, the final predicted risk training corpus includes region labels, enabling the subsequent model to learn based on the corresponding optical cable region during the learning process.

[0047] S5: Fine-tune the pre-trained semantic model according to the predicted risk training corpus corresponding to each optical cable area to obtain the fine-tuned semantic model. In this embodiment, the fine-tuning training specifically involves the pre-trained semantic model predicting the external damage risk of the corresponding optical cable area using the prediction risk training corpus. Specifically, the fine-tuning training process can be as follows: the pre-trained semantic model is fine-tuned using the corresponding prediction risk training corpus, so that the pre-trained semantic model can learn the optical cable external damage risk knowledge contained in the prediction risk training corpus, and predict the type and risk of external damage events based on the corresponding external damage event labels.

[0048] Meanwhile, as mentioned above, since the predicted risk training corpus corresponds to the optical cable area, the predicted risk training corpus corresponding to each optical cable area is used for fine-tuning training, so that the fine-tuned semantic model can predict the data of the corresponding optical cable area and obtain the risk of external damage events in the corresponding optical cable area.

[0049] In one implementation, obtaining the fine-tuned semantic model may include: The pre-trained semantic model is fine-tuned based on the prediction risk training corpus corresponding to each optical cable region, resulting in several fine-tuned semantic models. The number of fine-tuned semantic models matches the number of optical cable regions. In this embodiment, a fine-tuned semantic model is obtained for each optical cable region, and each fine-tuned semantic model can predict the risk of external damage events based solely on the data of its corresponding optical cable region.

[0050] S6: Obtain real-time external damage impact data and real-time acoustic data for each optical cable area; In this embodiment, the real-time acoustic data of the optical cable area can be obtained by extracting and restoring the optical cable data of the corresponding optical cable area.

[0051] In one implementation, a device equipped with a DAS (Distributed Acoustic Sensing) system can be connected to an optical cable and optical cable data from various optical cable areas can be collected. The optical cable data can be demodulated and restored to obtain the real-time acoustic data, which contains information features such as the frequency and intensity of the sound waves.

[0052] S7: The real-time external damage impact data is predicted using the fine-tuned semantic model to obtain several candidate external damage risk sources for the corresponding optical cable area.

[0053] Understandably, a knowledge graph is a data structure with data storage capabilities; without changing the structure of entities and relationships, only specific instances need to be added. Therefore, in this embodiment, as... Figure 5 As shown, the step of predicting the real-time external damage impact data using the fine-tuned semantic model to obtain several candidate external damage risk sources for the corresponding optical cable area may specifically include the following steps: S71: Based on the real-time external damage impact data and the real-time risk source information, generate specific instances of the corresponding entities in the corresponding local risk prediction knowledge graph, and update the corresponding local risk prediction knowledge graph based on the generated specific instances. As described above, since the local risk prediction knowledge graph has the ability to store data, for specific instances generated based on real-time external damage risk sources and real-time risk source information, it is not necessary to rebuild a new local risk prediction knowledge graph. The generated specific instances can be added to the local risk prediction knowledge graph of the corresponding optical cable area, or it can be checked whether a prior knowledge graph already exists in the corresponding local risk prediction knowledge graph. If it does, the corresponding specific instance will be updated.

[0054] Among them, a "join time" attribute can be added to the corresponding specific instance to distinguish between historical external damage impact data and real-time external damage impact data.

[0055] S72: Construct several real-time predictive semantic features based on specific instances of the updated local risk prediction knowledge graph; In this embodiment, the construction of the real-time prediction semantic features is similar to the construction of the external damage impact semantic features described above, and can be referred to the description of step S41 above, which will not be elaborated further here.

[0056] S73: The fine-tuned semantic model is used to predict several real-time predicted semantic features of the corresponding optical cable area to obtain several candidate external damage risk sources for the corresponding optical cable area.

[0057] Specifically, the real-time predicted semantic features of the corresponding optical cable area can be integrated and corresponding area labels can be added, so that the fine-tuned semantic model can determine the optical cable area corresponding to the real-time predicted semantic features, and then predict the candidate external damage risk sources of the optical cable based on the "area features" of the corresponding optical cable area.

[0058] In one implementation, the fine-tuned semantic model can be guided, or configured to generate output results in a fixed format, making it output understandable prediction text, such as "In optical cable area X, one of the candidate external damage risk sources is: the road at coordinates xx.xx is under construction, the distance from the optical cable is xx, and the probability of the optical cable breaking is xx." By having the fine-tuned semantic model output understandable prediction text, people can quickly and intuitively understand the probability of external damage and the causes of this risk.

[0059] In this embodiment, the finely tuned semantic model, after fine-tuning training, can effectively make targeted predictions for different regions by combining the potential correlation between external damage events among various real-time predicted semantic features with the "regional features" of the corresponding optical cable region, thereby effectively improving prediction accuracy. At the same time, by utilizing the principles of knowledge graphs and making predictions based on the semantic model, the risk prediction process and results become readable, improving the interpretability of the prediction results.

[0060] S8: Obtain corresponding candidate acoustic wave data based on the candidate external damage risk sources, compare each candidate acoustic wave data with the corresponding real-time acoustic wave data, and obtain the target external damage risk sources for each optical cable area.

[0061] In this embodiment, the candidate acoustic wave data can be obtained in advance by applying the same item or event as the candidate external damage risk source to the optical cable under the same or similar scenario, and then reconstructing the corresponding optical cable data collected by the optical cable.

[0062] Understandably, when candidate acoustic data matching the real-time acoustic data is found among the candidate acoustic data, it indicates that there is a risk source in the corresponding optical cable area that could lead to an external damage event, i.e., the target external damage risk source. In this embodiment, by first using the fine-tuned semantic model to obtain candidate external damage risk sources that may cause external damage events in each optical cable area, and by monitoring the optical cable data in each optical cable area, an attempt is made to find the target external damage risk source from the candidate external damage risk sources, thereby achieving accurate prediction of the source of the external damage event and thus effective prevention of the external damage event.

[0063] like Figure 6 As shown in the illustration, this application also provides a system for predicting the risk of external damage to optical cables. Optionally, the system for predicting the risk of external damage to optical cables may include: Knowledge graph ontology construction module 11 is used to construct a risk prediction knowledge graph ontology based on knowledge of optical cable external damage risk. In this embodiment, the knowledge graph ontology construction module 11 can be used to execute... Figure 1 For a detailed description of the knowledge graph ontology construction module 11 shown in step S1, please refer to the description of step S1.

[0064] Historical data acquisition module 12 is used to acquire historical external damage impact data for each area of ​​the optical cable based on knowledge of optical cable external damage risk; In this embodiment, the historical data acquisition module 12 can be used to perform... Figure 1 For a detailed description of the historical data acquisition module 12 shown in step S2, please refer to the description of step S2.

[0065] The local knowledge graph construction module 13 is used to construct a local risk prediction knowledge graph corresponding to each optical cable area based on the risk prediction knowledge graph ontology and according to the historical external damage impact data of each optical cable area. In this embodiment, the local knowledge graph construction module 13 can be used to execute Figure 1 For a detailed description of step S3, the local knowledge graph construction module 13 can be found in the description of step S3.

[0066] Training corpus generation module 14 is used to generate corresponding prediction risk training corpus based on the local risk prediction knowledge graph. In this embodiment, the training corpus generation module 14 can be used to perform... Figure 1 For a detailed description of the training corpus generation module 14 shown in step S4, please refer to the description of step S4.

[0067] The model training module 15 is used to fine-tune the pre-trained semantic model according to the prediction risk training corpus corresponding to each optical cable area, so as to obtain the fine-tuned semantic model. In this embodiment, the model training module 15 can be used to perform... Figure 1 For a detailed description of the model training module 15 shown in step S5, please refer to the description of step S5.

[0068] Real-time data acquisition module 16 is used to acquire real-time external damage impact data and real-time acoustic wave data of the optical cable area where the external damage event occurred; In this embodiment, the real-time data acquisition module 16 can be used to perform... Figure 1 For a detailed description of the real-time data acquisition module 16 shown in step S6, please refer to the description of step S6.

[0069] The candidate risk source acquisition module 17 is used to predict the real-time external damage impact data through the fine-tuned semantic model and acquire several candidate external damage risk sources for the corresponding optical cable area. In this embodiment, the candidate risk source acquisition module 17 can be used to perform... Figure 1 For a detailed description of the candidate risk source acquisition module 17 shown in step S7, please refer to the description of step S7. The target risk source acquisition module 18 is used to acquire corresponding candidate acoustic wave data based on the candidate external damage risk sources, compare each candidate acoustic wave data with the corresponding real-time acoustic wave data, and acquire the target external damage risk sources of each optical cable area.

[0070] In this embodiment, the candidate target risk source acquisition module 18 can be used to perform... Figure 1 For a detailed description of the target risk source acquisition module 18 shown in step S8, please refer to the description of step S8. This application provides an electronic device 20, the structure of which is as follows: Figure 7 As shown.

[0071] like Figure 7 As shown, the electronic device 20 includes a memory 21, a processor 22, a communication module 23, and an input / output interface 24, etc. Optionally, the memory 21, the processor 22, the communication module 23, and the input / output interface 24 can be connected and communicate with each other through a bus 25.

[0072] The memory 21 is used to store one or more computer programs and to transfer the code of the computer programs to the processor 22; when the one or more computer programs are executed by the processor 22, the optical cable external damage risk prediction method in this embodiment of the application is implemented.

[0073] Optionally, the electronic device 20 can be connected to a network via the communication module 23 to communicate with other devices, such as terminals or servers, to achieve data interaction. The electronic device 20 can be various forms of digital computers, exemplarily such as desktop computers, servers, workbenches, mainframes, or other types of computers. The electronic device 20 can also be various forms of mobile terminals, exemplarily such as smartphones, tablets, wearable devices (such as helmets, glasses, watches, etc.), and other similar mobile terminals.

[0074] Optionally, the electronic device 20 can connect to necessary input / output devices, such as a keyboard or display device, via the input / output interface 24. The electronic device 20 itself may have a display device, and other display devices can also be connected externally via the input / output interface 24. Optionally, a storage device, such as a hard disk, can also be connected via the input / output interface 24 to store data from the electronic device 20, read data from the storage device, or store data from the storage device in the memory 21. It is understood that the input / output interface 24 can be a wired interface or a wireless interface. Depending on the actual application scenario, the device connected to the input / output interface 24 can be a component of the electronic device 20 or an external device connected to the electronic device 20 when needed.

[0075] Optionally, the memory 21 may be a volatile memory and / or a non-volatile memory. The volatile memory may be a random access memory, etc., and the non-volatile memory may be a read-only memory, a programmable read-only memory, an erasable programmable read-only memory, an electrically erasable programmable read-only memory, or a flash memory, etc.

[0076] Optionally, the computer program stored in the processor 22 can be divided into one or more modules, which are stored in the memory 21 and executed by the processor 22 to perform the method provided in this embodiment. The one or more modules can be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the electronic device 20.

[0077] Optionally, the processor 22 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 22 include, but are not limited to, a central processing unit, a graphics processing unit, a digital signal processor, various special-purpose artificial intelligence computing chips, various processors running machine learning model algorithms, and can also be any suitable controller, microcontroller, processor, etc. The processor 22 executes the various methods and processes of this embodiment, exemplarily, such as a method for predicting the external damage risk of optical cables according to an embodiment of this application.

[0078] Optionally, the bus 25 may include a path for transmitting information. The bus 25 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Depending on its function, the bus 25 may be divided into an address bus, a data bus, a control bus, etc.

[0079] In an optional implementation, this application embodiment also provides a computer storage medium storing a computer program thereon. When the computer program is executed by a computer, it enables the computer to perform the methods described in the above-described method embodiments. Part or all of the computer program can be loaded and / or installed on the memory 21 of the electronic device 20. When the computer program is executed by the processor 22, one or more steps of a method for predicting the external damage risk of optical cables according to this application embodiment can be performed.

[0080] Optionally, the computer-readable storage medium may be a random access memory, a read-only memory, a programmable read-only memory, an erasable programmable read-only memory, an electrically erasable programmable read-only memory, etc.

[0081] Obviously, the above embodiments of this application are merely examples for clearly illustrating the technical solution of this application, and are not intended to limit the specific implementation of this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the claims of this application should be included within the protection scope of the claims of this application.

Claims

1. A method for predicting the risk of external damage to optical cables, characterized in that, The following steps are involved: Construct a risk prediction knowledge graph ontology based on knowledge of optical cable external damage risks; Historical data on the impact of external damage to optical cables in various regions of the optical cable are obtained based on knowledge of external damage risks. These regions are defined according to the pre-defined impact range of each optical cable segment. Based on the risk prediction knowledge graph ontology, a local risk prediction knowledge graph corresponding to each optical cable area is constructed according to the historical external damage impact data of each optical cable area. Generate corresponding prediction risk training corpus based on the local risk prediction knowledge graph; The pre-trained semantic model is fine-tuned based on the prediction risk training corpus corresponding to each optical cable area to obtain the fine-tuned semantic model; the fine-tuning training specifically involves the pre-trained semantic model using the prediction risk training corpus to predict external damage risk sources for the corresponding optical cable area. Acquire real-time data on the impact of external damage to each optical cable area and real-time acoustic data; The real-time acoustic data is obtained by extracting and restoring the optical cable data of the corresponding optical cable area; The real-time external damage impact data is predicted using the fine-tuned semantic model to obtain several candidate external damage risk sources for the corresponding optical cable area; Based on the candidate external damage risk sources, obtain the corresponding candidate acoustic wave data, compare each candidate acoustic wave data with the corresponding real-time acoustic wave data, and obtain the target external damage risk sources for each optical cable area.

2. The method for predicting the risk of external damage to optical cables according to claim 1, characterized in that, The specific process of constructing a risk prediction knowledge graph ontology based on knowledge of optical cable external damage risk is as follows: Knowledge extraction is performed on the knowledge of external damage risk of optical cable to obtain the entities and relationships of the knowledge of external damage risk of optical cable; Based on the entities and relationships, the structure of the risk prediction knowledge graph ontology is constructed to obtain the risk prediction knowledge graph ontology.

3. The method for predicting the risk of external damage to optical cables according to claim 1, characterized in that, The historical external damage impact data includes several historical external damage risk sources, as well as historical risk source information corresponding to the historical external damage risk sources. The historical external damage risk sources are divided into several risk types according to the risk prediction knowledge graph ontology. The real-time external damage impact data includes several real-time external damage risk sources, as well as real-time risk source information corresponding to the real-time external damage risk sources. The real-time external damage risk sources obtain their corresponding risk types based on the risk prediction knowledge graph ontology.

4. The method for predicting the risk of external damage to optical cables according to claim 3, characterized in that, Based on the aforementioned risk prediction knowledge graph ontology, a local risk prediction knowledge graph corresponding to each optical cable region is constructed according to the historical external damage impact data of each optical cable region. The specific content is as follows: Based on the entities in the risk prediction knowledge graph ontology, the historical external damage risk sources corresponding to the historical risk types are obtained from the historical external damage impact data of each optical cable area; Generate specific instances of corresponding entities based on the corresponding historical external risk sources and historical risk source information; Based on the specific instances and corresponding entities, a local risk prediction knowledge graph for the corresponding optical cable area is constructed.

5. The method for predicting the risk of external damage to optical cables according to claim 3, characterized in that, The process involves using the fine-tuned semantic model to predict the real-time external damage impact data and obtain several candidate external damage risk sources for the corresponding optical cable area. The specific details are as follows: Based on the real-time external damage impact data and the real-time risk source information, a specific instance of the corresponding entity in the corresponding local risk prediction knowledge graph is generated, and the corresponding local risk prediction knowledge graph is updated based on the generated specific instance. Based on specific instances of the updated local risk prediction knowledge graph, several real-time prediction semantic features are constructed. The fine-tuned semantic model is used to predict several real-time predictive semantic features of the corresponding optical cable area to obtain several candidate external damage risk sources for the corresponding optical cable area.

6. The method for predicting the risk of external damage to optical cables according to claim 1, characterized in that, The step of generating corresponding prediction risk training corpus based on the local risk prediction knowledge graph specifically includes: Based on each specific instance in the local risk prediction knowledge graph, construct corresponding semantic features of external damage impact; Based on the relationships between entities corresponding to specific instances in the local risk prediction knowledge graph, the semantic features of external damage impact are integrated to obtain several sets of semantic features of external damage impact. Add corresponding external damage event tags to each of the aforementioned sets of semantic features related to external damage impact; Based on the semantic feature set of external damage impact after adding external damage event tags, construct the prediction risk training corpus corresponding to each local risk prediction knowledge graph, and add corresponding region tags to each prediction risk training corpus.

7. The method for predicting the risk of external damage to optical cables according to any one of claims 1-5, characterized in that, The step of fine-tuning the pre-trained semantic model based on the predicted risk training corpus corresponding to each optical cable region to obtain the fine-tuned semantic model specifically includes: The pre-trained semantic model is fine-tuned based on the predicted risk training corpus corresponding to each optical cable area, and several fine-tuned semantic models are obtained through the fine-tuning training; the number of fine-tuned semantic models matches the number of optical cable areas.

8. A system for predicting the external damage risk of optical cables, equipped with the method for predicting the external damage risk of optical cables according to claim 1, characterized in that, The system includes: The knowledge graph ontology construction module is used to construct a risk prediction knowledge graph ontology based on knowledge of optical cable external damage risks. The historical data acquisition module is used to acquire historical external damage impact data for each optical cable area on the optical cable based on the knowledge of external damage risk; the optical cable area is divided according to the influence range of each preset optical cable segment on the optical cable; The local knowledge graph construction module is used to construct a local risk prediction knowledge graph corresponding to each optical cable area based on the risk prediction knowledge graph ontology and the historical external damage impact data of each optical cable area. The training corpus generation module is used to generate corresponding prediction risk training corpus based on the local risk prediction knowledge graph. The model training module is used to fine-tune the pre-trained semantic model according to the prediction risk training corpus corresponding to each optical cable area, so as to obtain the fine-tuned semantic model; the fine-tuning training specifically refers to the pre-trained semantic model using the prediction risk training corpus to predict the external damage risk sources of the corresponding optical cable area. The real-time data acquisition module is used to acquire real-time external damage impact data and real-time acoustic wave data of the optical cable area where the external damage event occurred; the real-time acoustic wave data is extracted and restored based on the optical cable data of the corresponding optical cable area. The candidate risk source acquisition module is used to predict the real-time external damage impact data through the fine-tuned semantic model and acquire several candidate external damage risk sources for the corresponding optical cable area. The target risk source acquisition module is used to acquire corresponding candidate acoustic wave data based on the candidate external damage risk sources, compare each candidate acoustic wave data with the corresponding real-time acoustic wave data, and acquire the target external damage risk sources for each optical cable area.

9. An electronic device, characterized in that, include: Memory, used to store one or more computer programs; A processor, when the one or more computer programs are executed by the processor, implements the optical cable external damage risk prediction method as described in any one of claims 1-6.

10. A computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to cause the processor to implement the optical cable external damage risk prediction method as described in any one of claims 1-6.

Images