Communication cable routing detection method and system based on digital twinning

Abstract

This application provides a method and system for detecting communication optical cable routes based on digital twins. The method acquires the marked location information in the communication optical cable and integrates multi-source information based on a differentiated spatiotemporal calibration algorithm. It constructs a digital twin model comprising a base layer, a dynamic layer, and an interaction layer. In the base layer, a composite framework including spatial topological relationships, environmental constraint boundaries, and physical characteristic benchmarks is formed. In the dynamic layer, a knowledge graph and rule engine are introduced to perform semantic parsing of dynamic data. Simultaneously, a correlation reasoning model between dynamic data and environmental features in the base layer is constructed. When data anomalies are detected, the potential correlation between the anomalies and environmental features is mined, and the model's sensitivity to different data is dynamically adjusted based on the reasoning results. In the interaction layer, an operation and maintenance scenario feature library is constructed, and scenario adaptation and pre-feedback mechanisms are introduced. This detection method significantly improves the accuracy of detecting communication optical cable routes in complex environments.

Description

Technical Field

[0001] The embodiments of this application belong to the field of communication technology, and more specifically, relate to a communication optical cable route detection method and system based on digital twins. Background Technology

[0002] In the field of optical communication network operation and maintenance, optical cables, as the core carrier of information transmission, are directly related to the smoothness and reliability of the communication network. With the rapid development of technologies such as 5G, cloud computing, and big data, the scale of optical communication networks continues to expand, and optical cable routes are becoming increasingly complex, covering not only urban underground and building-to-building areas, but also extending to remote mountainous areas, rivers, lakes, and seas. In operation and maintenance work, timely and accurate information on optical cable routes is crucial for fault location, daily inspections, line maintenance, and network expansion, and is the foundation for ensuring the efficient operation of optical communication networks.

[0003] Currently, the commonly used optical cable route detection technologies in optical communication network operation and maintenance mainly include the following: First, detection based on optical time domain reflectometer (OTDR), which analyzes the transmission and reflection of light pulses in the optical cable to obtain the length, loss, and approximate location of the fault point of the optical cable, assisting operation and maintenance personnel in judging the line status; Second, relying on the routing records of geographic information system (GIS), the optical cable route information is pre-entered into the system, and the route direction is understood by querying GIS data during operation and maintenance; Third, manual inspection combined with marker identification, where operation and maintenance personnel carry positioning equipment to the site, judge the optical cable route based on ground markings or experience, and record relevant information.

[0004] However, these existing technologies have significant shortcomings in actual optical communication network operation and maintenance. While OTDR-based detection can locate fault points, it cannot intuitively present the actual route of the optical cable, especially in areas with densely laid optical cables or multiple rerouting, making it difficult to meet the demand for accurate route paths in operation and maintenance. Routing records relying on GIS are often not updated in a timely manner due to line modifications and construction changes, resulting in discrepancies between the routing information obtained by operation and maintenance personnel and the actual situation, affecting maintenance efficiency. Manual inspection combined with marker identification is limited by complex terrain, missing or blurred markers, etc., which is not only time-consuming and labor-intensive, but also difficult to accurately detect hidden routes such as those buried underground or passing through buildings, which is prone to misjudgment and hinders the daily maintenance and emergency repair of optical cables. Summary of the Invention

[0005] The purpose of this application is to provide a method and system for detecting optical fiber routing based on digital twins, so as to at least solve the technical problems of low detection efficiency, inaccurate detection of hidden parts, and easy misjudgment in related optical fiber routing detection technologies.

[0006] To achieve the above objectives, the embodiments of this application provide the following technical solutions.

[0007] According to one embodiment of this application, a communication optical cable route detection method based on digital twin is provided, comprising the following steps:

[0008] Obtain the location information of markers in communication optical cables and integrate multi-source information based on a differentiated spatiotemporal calibration algorithm;

[0009] A digital twin model is constructed, comprising a foundation layer, a dynamic layer, and an interaction layer. In the foundation layer, a multi-dimensional foundational model with dual mappings is built. Based on the mapping of optical cable physical parameters to geographic coordinates, the inherent characteristics of the environment along the route are incorporated, forming a composite foundational framework that includes spatial topological relationships, environmental constraint boundaries, and physical characteristic benchmarks. In the dynamic layer, after real-time access to multi-source dynamic data, a knowledge graph and rule engine are introduced to perform semantic parsing of the dynamic data. Simultaneously, a correlation reasoning model between dynamic data and environmental features in the foundation layer is constructed. When data anomalies are detected, the potential correlation between the anomalies and environmental features is explored, and the model's sensitivity to different data is dynamically adjusted based on the reasoning results. In the interaction layer, an operation and maintenance scenario feature library is constructed, and scenario adaptation and pre-feedback mechanisms are introduced.

[0010] Based on the visualization elements generated by the interaction layer, the virtual model of the optical cable route is overlaid on the physical scene to perform detection operations on the target route segment. The detection data is fed back to the dynamic layer of the digital twin model in real time to update the status of the virtual model. The dynamic layer of the digital twin model is used to analyze multi-source dynamic data. Combined with the composite basic framework of the basic layer, a route status evaluation system is constructed, and abnormal status of the route is identified through a preset threshold judgment mechanism.

[0011] Preferably, the multi-source information includes OTDR light pulse reflection data, GIS geographic coordinate data, fiber optic distributed sensing data, and manual inspection records;

[0012] The step of integrating multi-source information based on the differentiated spatiotemporal calibration algorithm includes spatial calibration and temporal calibration. The spatially calibrated multi-source data is superimposed and displayed in a three-dimensional coordinate system. The calibration effect is verified by calculating the spatial distance deviation between data points and the correlation coefficient of the time series, until the deviation does not exceed a threshold. Wherein:

[0013] In spatial calibration, for OTDR optical pulse reflection data, an optical path-to-actual distance conversion model is established based on the physical parameters of the optical cable. The optical path data calculated by the optical pulse transmission time is converted into the actual geographical distance in meters. Combining the topological relationship of the GIS vector map, the distance data of the OTDR is anchored to the two-dimensional coordinate system of the GIS through a coordinate mapping algorithm to achieve spatial alignment. For fiber optic grating distributed sensing data, the GPS coordinates of the sensor deployment location are extracted and matched with the optical cable routing nodes in the GIS map. The spatial offset caused by sensor installation errors is corrected by the triangulation algorithm, so that the sensing data corresponds one-to-one with the optical cable routing location. The unstructured location descriptions in the manual inspection records are parsed into geographical coordinates through natural language processing technology and then compared and calibrated with the GIS data.

[0014] In time calibration, the Dynamic Time Warping (DTW) algorithm is used to stretch or compress the time axis of time-series data to address the differences in sampling frequencies of different devices; the time information of manual inspection records is aligned with the time axis through time standardization processing.

[0015] Preferably, the step of introducing a biological mimicry evolutionary mechanism to form a new routing model adapted to the current environment in the digital twin model includes:

[0016] Data were collected on typical routing characteristics of optical cables under different environments. Principal component analysis was used to map environmental parameters and routing characteristics into multidimensional feature vectors to construct a routing gene library for communication optical cables.

[0017] When the digital twin model receives new detection data or environmental information, it retrieves the corresponding feature parameter set as the initial solution.

[0018] A genetic algorithm is used to perform crossover and mutation operations on the retrieved feature parameters. Combined with the mechanical properties of optical cable materials, physical engine constraints are constructed to simulate the stress and deformation process of optical cable under new environment and generate multiple sets of candidate route forms.

[0019] By optimizing multiple objectives such as minimizing transmission loss, maximizing mechanical stability, and minimizing maintenance costs, the optimal solution is selected from the candidate model set to form a new routing model that adapts to the current environment.

[0020] Preferably, the step of generating multiple sets of candidate route patterns includes:

[0021] The optical cable routing pattern is decomposed into quantifiable genetic parameters, and a genetic chain is formed by two-dimensional encoding of parameters and thresholds.

[0022] Historical routing parameters with a matching degree greater than a threshold for the current environment characteristics are retrieved from the routing gene library, and several initial routing schemes are generated by random combination.

[0023] Import the mechanical parameters of the optical cable material and environmental constraints to construct a finite element analysis model;

[0024] For each routing scheme in the population, its performance in the actual environment is simulated through the physics engine, multiple indicators are output and the fitness value is calculated by weighting, and the routing schemes whose fitness values ​​meet the requirements are retained.

[0025] For the retained routing schemes, two groups are randomly selected as parents, and gene fragments are exchanged in a segmented crossover manner. The gene parameters of the offspring are randomly modified with a preset probability to simulate gene mutations in biological evolution.

[0026] The offspring routing schemes generated by crossover and mutation are verified, and the process is iterated based on the verification results until a new routing form is obtained.

[0027] Preferably, the method further includes a step of detecting the hidden segment route, including:

[0028] Based on the digital twin model, a routing feature transfer learning module is introduced. The features of non-hidden segments are used as basic samples and transferred to the hidden segment sub-model of the digital twin model through federated learning. Combined with the indirect feature data around the hidden segments collected by the dynamic layer, an implicit mapping model of the environmental routing state is trained and generated.

[0029] A multimodal data coupling network is constructed based on a digital twin model. The feature vectors of vibration, optical signals and geological structures are weighted through an attention mechanism to generate a dynamically updated three-dimensional routing model for hidden segments.

[0030] Based on the three-dimensional routing model of the concealed segment, the state evolution of the concealed segment under environmental stress is simulated through a digital twin physical engine, and high-risk areas are marked.

[0031] Preferably, in the digital twin model, the step of introducing a biological mimicry evolutionary mechanism to form a new routing model form adapted to the current environment further includes:

[0032] For concealed route segments, an inverse mathematical model is established using known endpoint parameters and ground-penetrating radar point cloud data;

[0033] By minimizing the objective function of routing curvature and material stress, the most likely continuous routing trajectory is derived and incorporated into the candidate morphology set.

[0034] Preferably, the steps in the interaction layer to construct an operation and maintenance scenario feature library and introduce scenario adaptation and pre-feedback mechanisms include:

[0035] The operation and maintenance scenario is decomposed into a triplet feature vector, and the mapping relationship between the scenario and multi-source data is established through knowledge graph;

[0036] Based on historical operation and maintenance records, a reinforcement learning algorithm is used to dynamically adjust the weights of each parameter in the feature library;

[0037] The concept of scenario lifecycle is introduced to construct a scenario evolution state machine, with each state corresponding to specific interaction rules and data priority configurations. On-site environmental information is collected by sensors and combined with real-time data from the digital twin model. A multimodal fusion algorithm is used to identify the current operation and maintenance scenario. Based on the identified scenario, the corresponding parameter configuration is retrieved from the feature library, and the optimal set of interaction rules is generated through a genetic algorithm. The interactive interface is dynamically reconstructed according to the hardware parameters of the interactive device and the current scenario requirements.

[0038] Before the user performs a detection operation, the physical engine based on the digital twin model, combined with the current scene characteristics, simulates the possible results of the operation and establishes a pre-feedback decision tree to conduct a risk assessment on the simulation results; the actual operation results of the user are compared with the pre-feedback predictions, and the pre-feedback model parameters are optimized through meta-learning algorithms.

[0039] Preferably, in the dynamic layer of the digital twin model:

[0040] Dynamic data and basic layer environmental features are mapped to entity nodes in a knowledge graph. Semantic relationships between entities are extracted using sequence labeling. A hierarchical structure containing four types of ontology—physical entities, environmental elements, event types, and indicator parameters—is constructed. Logical relationships between ontology are defined. Multi-source heterogeneous data are fused using DS evidence theory.

[0041] The design incorporates a three-tiered rule system comprising domain rules, association rules, and inference rules, with each rule configured with a weight coefficient and a trigger threshold. A forward chain inference algorithm is employed to match real-time data with entities and relationships in the knowledge graph, transforming raw data into interpretable event tags and constructing a three-tiered mapping of data, features, and events.

[0042] A graph neural network that constructs a spatiotemporal attention mechanism integrates the temporal and spatial dimensions into associative reasoning, represented as follows:

[0043]

[0044] In the formula, Let v represent the feature vector of node v at time t. Represents the set of adjacent nodes. Represents the spatiotemporal attention weights. , Represents the learnable weight matrix; This represents the activation function. This represents the eigenvector of node u at time t;

[0045] We employ PC algorithm combined with Granger causality test to explore the causal relationship between dynamic data and environmental features;

[0046] Anomaly detection is performed using the Isolation Forest algorithm combined with knowledge graph constraints. When an anomaly is detected, the knowledge graph is queried through a rule engine to extract related environmental features, and a formula for calculating the correlation degree of environmental anomalies is constructed, expressed as: Correlation degree = λ1S KG +λ2S ST +λ3S C Among them, S KG For the semantic similarity of knowledge graphs, S ST S represents the spatiotemporal co-occurrence frequency. C λ1, λ2, and λ3 represent the causal strength, and λ3 are weight coefficients that are dynamically adjusted through reinforcement learning. Based on the correlation evaluation results, the sensitivity of the model to different data is dynamically adjusted.

[0047] Preferably, in the step of extracting semantic relationships between entities using sequence labeling, a sequence labeling model that integrates dynamic attention mechanism and knowledge graph constraints is used to extract entity semantic relationships, wherein the sequence labeling model is represented as:

[0048]

[0049] In the formula, x represents the input text sequence. Used to capture contextual information at the i-th position of the sequence. This represents the attention weight at position i calculated through the attention mechanism, used to dynamically adjust the weights at different positions. CRF stands for Conditional Random Field Function, used to handle dependencies between labels.

[0050] Loss function L with knowledge graph constraints KG It is represented by cosine similarity, where the loss function L KG Represented as:

[0051]

[0052] In the formula, For the predicted relation vector, This represents the corresponding relation vector in the knowledge graph.

[0053] According to another embodiment of this application, a communication optical cable routing detection system based on digital twins is provided, comprising the following modules:

[0054] The information integration module is used to acquire the marker location information in the communication optical cable and integrate multi-source information based on the differentiated spatiotemporal calibration algorithm;

[0055] The twin model construction module is used to build a digital twin model comprising a base layer, a dynamic layer, and an interaction layer. In the base layer, a multi-dimensional base model with dual mappings is constructed. Based on the mapping of optical cable physical parameters to geographic coordinates, the inherent characteristics of the environment along the route are incorporated to form a composite base framework including spatial topological relationships, environmental constraint boundaries, and physical characteristic benchmarks. In the dynamic layer, after real-time access to multi-source dynamic data, a knowledge graph and rule engine are introduced to perform semantic parsing of the dynamic data. Simultaneously, a correlation reasoning model between dynamic data and environmental features in the base layer is constructed. When data anomalies are detected, the potential correlation between the anomalies and environmental features is mined, and the model's sensitivity to different data is dynamically adjusted based on the reasoning results. In the interaction layer, an operation and maintenance scenario feature library is constructed, and scenario adaptation and pre-feedback mechanisms are introduced.

[0056] The route detection module is used to overlay a virtual model of the optical cable route onto the physical scene based on the visualization elements generated by the interaction layer, perform detection operations on the target route segment, and feed the detection data back to the dynamic layer of the digital twin model in real time to update the status of the virtual model. The dynamic layer of the digital twin model is used to analyze multi-source dynamic data, and combined with the composite basic framework of the basic layer, a route status evaluation system is constructed. An abnormal state of the route is identified through a preset threshold judgment mechanism.

[0057] Compared with existing technologies, the beneficial effects of the digital twin-based communication optical cable route detection method and system of this application are:

[0058] First, the digital twin model constructed in this application deeply integrates the physical parameters of the optical cable with the inherent characteristics of the environment, solving the problem that traditional models only focus on the parameters of the optical cable itself and ignore environmental constraints, thus providing a more comprehensive benchmark framework for detection. The knowledge graph and association reasoning mechanism of the dynamic layer of the model break through the limitation of a single data dimension, and establish causal relationships between dynamic data and environmental features through semantic parsing, automatically identifying potential relationships and making it more predictive. The interaction layer, by constructing an operation and maintenance scenario feature library and introducing scenario adaptation and pre-feedback mechanisms, realizes scenario-based adaptation of detection interaction and proactive avoidance of operational risks.

[0059] Second, the dynamic layer's associative reasoning model in this application combines the semantic constraints of the knowledge graph and verifies the validity of the data through multi-dimensional feature coupling, thereby reducing the false alarm rate; the dynamic layer's mechanism for adjusting data sensitivity based on reasoning results enables the model to adapt to environmental changes.

[0060] Third, this application uses interactive layer visualization elements to overlay virtual models and physical scenes, enabling operation and maintenance personnel to intuitively obtain routing information that is isomorphic between the virtual and physical worlds, thus avoiding missed or false detections. In addition, the physical detection data of this application is synchronously transmitted back to the dynamic layer, triggering real-time updates to the virtual model, ensuring that the model always maintains consistency with the physical routing. Furthermore, an evaluation system is constructed in conjunction with the basic layer composite framework, and a threshold judgment mechanism is used to achieve multi-dimensional verification of abnormal states, significantly improving the accuracy and reliability of routing detection. Attached Figure Description

[0061] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0062] In the attached diagram:

[0063] Figure 1 This is a flowchart illustrating the implementation of the communication optical cable route detection method based on digital twins in this application.

[0064] Figure 2 This is a structural block diagram of a digital twin model provided in an embodiment of the present invention;

[0065] Figure 3 This is a sub-flowchart of the communication optical cable route detection method based on digital twins, as described in this application.

[0066] Figure 4 This is another sub-flowchart of the communication optical cable route detection method based on digital twins in this application embodiment;

[0067] Figure 5 This is a structural block diagram of a communication optical cable routing detection system based on digital twins, as described in an embodiment of this application.

[0068] Figure 6 This is a hardware structure block diagram of a computer terminal for a communication optical cable route detection method based on digital twins according to an embodiment of this application. Detailed Implementation

[0069] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0070] It should be noted that, unless otherwise specified, all technical and scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0071] In this invention, unless otherwise stated, directional terms such as "upper," "lower," "top," and "bottom" are generally used in relation to the direction shown in the accompanying drawings, or in relation to the vertical, perpendicular, or gravitational direction of the component itself; similarly, for ease of understanding and description, "inner" and "outer" refer to the inner and outer contours of each component itself, but the above directional terms are not intended to limit this invention.

[0072] According to one embodiment of this application, a communication optical cable route detection method based on digital twin is provided;

[0073] This route detection method aims to at least address the technical problems of low detection efficiency, inaccurate detection of concealed parts, and easy misjudgment in related communication optical cable route detection technologies.

[0074] Please refer to Figure 1 and Figure 2 This application provides a communication optical cable route detection method based on digital twin. In the communication optical cable route detection method, the constructed digital twin model includes a base layer 201, a dynamic layer 202, and an interaction layer 203.

[0075] Based on the digital twin model constructed above, the communication optical cable route detection method of this application includes the following steps:

[0076] S101: Obtain the marker location information in the communication optical cable and integrate multi-source information based on the differentiated spatiotemporal calibration algorithm;

[0077] In the specific implementation of step S101 of this application, the marked location information is the basis for constructing a precise mapping of the digital twin model. Multi-source information (such as OTDR optical path data, GIS coordinates, and sensor data) comes from different devices. By using a unified location benchmark, it is ensured that the multi-source data are aligned in the same coordinate system. The marked location information can be selected from key nodes of the optical cable route as marked points, including inherent physical nodes and manually added marks. Inherent physical nodes include optical cable junction boxes, terminal boxes, branch points, bends, and points of change in burial depth. Manually added marks include deploying passive RFID tags or fiber optic grating marks at intervals of 50-100 meters on long straight sections without inherent nodes, and recording location information through physical identification.

[0078] In some embodiments of this application, the multi-source information includes OTDR light pulse reflection data, GIS geographic coordinate data, fiber optic distributed sensing data, and manual inspection records;

[0079] Specifically, in some embodiments, the step of integrating multi-source information based on the differentiated spatiotemporal calibration algorithm includes spatial calibration and temporal calibration. The spatially calibrated multi-source data is superimposed and displayed in a three-dimensional coordinate system. The calibration effect is verified by calculating the spatial distance deviation between data points and the correlation coefficient of the time series until the deviation does not exceed the threshold.

[0080] Furthermore, in the spatial calibration provided in this embodiment, for OTDR optical pulse reflection data, an optical path-to-actual distance conversion model is established based on the physical parameters of the optical cable, converting the optical path data calculated by the optical pulse transmission time into the actual geographical distance in meters; combined with the topological relationship of the GIS vector map, the distance data of the OTDR is anchored to the two-dimensional coordinate system of the GIS through a coordinate mapping algorithm to achieve spatial alignment; for fiber optic grating distributed sensing data, the GPS coordinates of the sensor deployment location are extracted and matched with the optical cable routing nodes in the GIS map, and the spatial offset caused by sensor installation errors is corrected through a triangulation algorithm, so that the sensing data corresponds one-to-one with the optical cable routing location; the unstructured location description in the manual inspection record is parsed into geographical coordinates through natural language processing technology, and then compared and calibrated with the GIS data;

[0081] Furthermore, in the time calibration provided in this embodiment, the Dynamic Time Warping (DTW) algorithm is used to stretch or compress the time axis of the time-series data to address the differences in sampling frequencies of different devices; the time information recorded by manual inspection is aligned with the time axis through time standardization processing.

[0082] The communication optical cable route detection method in this application embodiment further includes step S102, in which a digital twin model including a base layer, a dynamic layer and an interaction layer is constructed.

[0083] In the foundational layer, a multi-dimensional foundational model with dual mappings is constructed. Based on the mapping of optical cable physical parameters and geographic coordinates, the inherent characteristics of the environment along the route are incorporated to form a composite foundational framework that includes spatial topological relationships, environmental constraint boundaries, and physical characteristic benchmarks. Among these, the inherent characteristics of the environment along the route include geology, hydrology, surrounding facilities, and climate conditions; spatial topological relationships include node connection relationships and spatial hierarchical relationships, which are stored using the graph database Neo4j; in addition, environmental constraint boundaries are defined based on environmental characteristics to define the safe operation boundaries of the optical cable. By integrating mapping relationships and environmental characteristics, the foundational layer forms a composite foundational framework containing three major elements, providing a static benchmark for the digital twin model.

[0084] In the digital twin model constructed in this application, the physical parameters of the optical cable are deeply integrated with the inherent characteristics of the environment, which solves the problem that the traditional model only focuses on the parameters of the optical cable itself and ignores environmental constraints, and provides a more comprehensive benchmark framework for detection.

[0085] In the dynamic layer, after real-time access to multi-source dynamic data, knowledge graphs and rule engines are introduced to perform semantic parsing of dynamic data. At the same time, a correlation reasoning model between dynamic data and environmental features of the basic layer is constructed. When data anomalies are detected, the potential correlation between anomalies and environmental features is explored, and the sensitivity of the model to different data is dynamically adjusted based on the reasoning results.

[0086] In this embodiment, the semantic parsing of multi-source dynamic data is achieved through the entity relationship framework provided by the knowledge graph and the logical reasoning capability of the rule engine. Specifically, a knowledge graph containing dynamic data entities, environmental entities, event types, and their relationships is pre-constructed. A rule base is built based on domain experience, including basic rules, association rules, and reasoning rules. These rules are invoked to match the pre-processed dynamic data and output semantic tags with confidence. The parsing results are then synchronously updated to the knowledge graph. New entity relationships are extracted through sequence labeling algorithms to improve the accuracy of the parsing.

[0087] In the reasoning model, reasoning is achieved through a collaborative mechanism of spatial association modeling, temporal causal verification, and semantic constraint optimization. In the spatial dimension, a constructed knowledge graph serves as the framework, abstracting dynamic data entities and environmental feature entities into graph nodes. A graph neural network (GNN) is used to learn spatial dependencies between nodes, enabling the model to focus on spatial relationships. In the temporal dimension, the time series of dynamic data with sequential associations is determined. Granger causality tests are used to verify whether changes in environmental features precede anomalies in dynamic data, providing temporal causal evidence for association reasoning. In the semantic dimension, knowledge graph constraints reduce noise. Pre-defined relationships within the knowledge graph constrain the reasoning process, preventing irrelevant data from being misjudged as related data and improving the robustness of reasoning.

[0088] The dynamic layer's associative reasoning model in this application combines the semantic constraints of a knowledge graph and verifies data validity through multi-dimensional feature coupling, reducing the false alarm rate. The dynamic layer's mechanism for adjusting data sensitivity based on reasoning results enables the model to adapt to environmental changes. Therefore, the knowledge graph and associative reasoning mechanism of the dynamic layer in this application's embodiment model break through the limitations of a single data dimension and establishes causal relationships between dynamic data and environmental features through semantic parsing, automatically identifying potential relationships and making it more predictive.

[0089] In the interaction layer, an operation and maintenance scenario feature library is constructed, and a scenario adaptation and pre-feedback mechanism is introduced. Specifically, the operation and maintenance scenario is decomposed into triple feature vectors, and a mapping relationship between the scenario and multi-source data is established through a knowledge graph. Based on historical operation and maintenance records, a reinforcement learning algorithm is used to dynamically adjust the weights of each parameter in the feature library. The parameters in the operation and maintenance scenario feature library refer to the key features describing the scenario, such as fault type, environmental interference intensity, data priority, etc., while the weights represent the importance of these parameters in scenario decision-making. The weights are optimized through reinforcement learning, and finally, the parameter weights are dynamically adapted to the scenario.

[0090] Furthermore, in this embodiment, the concept of a scenario lifecycle is introduced, and a scenario evolution state machine is constructed. Each state corresponds to specific interaction rules and data priority configurations. The state machine is a model used to describe the state transition rules. When the fault repair scenario switches from the fault location state (OTDR data analysis should be prioritized) to the on-site repair state, the triggering condition may be determining the fault location coordinates. The state machine automatically completes the state transition through preset rules to ensure that the scenario stage is synchronized with the actual operation rhythm. For example, if there are no new alarms within 3 minutes after the fault is repaired, it switches to the recovery verification state.

[0091] This application embodiment collects on-site environmental information through sensors, combines it with real-time data from a digital twin model, uses a multimodal fusion algorithm to identify the current operation and maintenance scenario, retrieves corresponding parameter configurations from a feature library based on the identified scenario, and generates an optimal set of interaction rules through a genetic algorithm;

[0092] Furthermore, based on the hardware parameters of the interactive device and the current scenario requirements, a responsive design algorithm is used to dynamically reconstruct the interactive interface; before the user performs a detection operation, based on the physical engine of the digital twin model and combined with the characteristics of the current scenario, the possible results of the operation are simulated, and a pre-feedback decision tree is established to conduct a risk assessment on the simulation results; the actual operation results of the user are compared with the pre-feedback predictions, and the pre-feedback model parameters are optimized through a meta-learning algorithm.

[0093] The interaction layer of this application realizes the scenario-based adaptation of detection interaction and the proactive avoidance of operational risks by constructing an operation and maintenance scenario feature library and introducing scenario adaptation and pre-feedback mechanisms.

[0094] Please continue to refer to Figure 1 The communication optical cable route detection method of this application embodiment further includes step S103:

[0095] Based on the visualization elements generated by the interaction layer, a virtual model of the optical cable route is overlaid on the physical scene to perform detection operations on the target route segment. The detection data is fed back to the dynamic layer of the digital twin model in real time to update the status of the virtual model. The dynamic layer of the digital twin model is used to analyze multi-source dynamic data, and combined with the composite basic framework of the basic layer, a route status evaluation system is constructed. Among them, an abnormal status of the route is identified through a preset threshold judgment mechanism.

[0096] In step S103, the interaction layer generates visualization elements adapted to the physical scene based on the spatial topology of the base layer and the real-time status of the dynamic layer. These elements include a three-dimensional virtual routing model and key status indicators. Maintenance personnel can intuitively see the virtual model superimposed on the physical route through AR glasses or terminal devices and quickly locate the target detection segment.

[0097] In addition, the constructed routing state evaluation system includes physical performance indicators, environmental adaptability indicators, and historical comparison indicators. Among them, physical performance indicators are based on the physical characteristics benchmark of the base layer to evaluate optical signal quality and mechanical stability; environmental adaptability indicators are combined with the environmental constraint boundaries of the base layer to evaluate the adaptability of the routing segment to the environment; and historical comparison indicators are evaluated through time-series data analysis of the dynamic layer to assess the trend of state change.

[0098] This application achieves accurate overlay of virtual models and physical scenes through interactive layer visualization elements, allowing operation and maintenance personnel to intuitively obtain routing information that is isomorphic between the virtual and physical worlds, avoiding missed or false detections. In addition, the physical detection data of this application is synchronously transmitted back to the dynamic layer, triggering real-time updates to the virtual model, ensuring that the model always maintains consistency with the physical routes. Furthermore, an evaluation system is constructed in conjunction with the basic layer composite framework. This embodiment achieves multi-dimensional verification of abnormal states through a threshold judgment mechanism, significantly improving the accuracy and reliability of route detection.

[0099] like Figure 3 As shown, in one implementation of this application, the step of introducing a biological mimicry evolution mechanism into the digital twin model to form a new routing model form adapted to the current environment includes:

[0100] S301: Collect data on typical routing characteristics of optical cables under different environments, and map environmental parameters and routing characteristics into multi-dimensional feature vectors through principal component analysis to construct a communication optical cable routing gene library;

[0101] Among them, the routing gene library is a structured database formed by genetically encoding the routing characteristics of communication optical cables in different environments. It is used to transform complex routing characteristics and environmental parameters into quantifiable and heritable gene parameters, analogous to the genetic variation characteristics of biological genes, to provide basic gene material for the dynamic optimization of routing models.

[0102] S302: When the digital twin model receives new detection data or environmental information, it retrieves the corresponding feature parameter set as the initial solution;

[0103] S303: A genetic algorithm is used to perform crossover and mutation operations on the retrieved feature parameters. Combined with the mechanical properties of optical cable materials, physical engine constraints are constructed to simulate the stress and deformation process of optical cable under new environment and generate multiple sets of candidate route forms.

[0104] In some embodiments, physics engine constraints include stress constraints, bending radius constraints, and shear resistance constraints;

[0105] S304: Through multi-objective optimization that minimizes transmission loss, maximizes mechanical stability, and minimizes maintenance costs, the optimal solution is selected from the candidate form set to form a new routing model form that adapts to the current environment.

[0106] In this embodiment, in minimizing transmission loss, transmission loss includes optical cable bending loss, connector loss, and material attenuation; in maximizing mechanical stability, stability is measured by a safety factor, and the higher the safety factor, the stronger the stability; in minimizing maintenance cost, maintenance cost includes construction cost, inspection cost, and fault repair cost.

[0107] In the comprehensive evaluation of multi-objective optimization, a weighted summation method is used to integrate the three objectives into a single evaluation index F. The weight coefficients w1, w2, and w3 are dynamically adjusted through reinforcement learning, with w1 + w2 + w3 = 1. The evaluation index for multi-objective optimization is expressed as:

[0108]

[0109] In the formula, , and These respectively represent minimizing transmission loss, maximizing mechanical stability, and minimizing maintenance costs; , , This represents the maximum value of the three optimization objectives in the candidate set;

[0110] The optimization objective is to minimize F, that is: ;in, , and This represents the normalized target value;

[0111] The solution with the smallest F value is selected from the candidate route pattern set as the optimal solution. If there are multiple solutions with similar F values ​​(e.g., difference ≤ 0.05), further screening is carried out using the Pareto optimality criterion, including: prioritizing the solution that performs better on at least one objective (e.g., lower transmission loss or higher stability), and finally forming a new route model pattern that adapts to the current environment.

[0112] Furthermore, in this embodiment of the application, the step of generating multiple sets of candidate route patterns includes:

[0113] The optical cable routing pattern is decomposed into quantifiable genetic parameters, and a genetic chain is formed by two-dimensional encoding of parameters and thresholds.

[0114] Historical routing parameters with a matching degree greater than a threshold for the current environment characteristics are retrieved from the routing gene library, and several initial routing schemes are generated by random combination.

[0115] Import the mechanical parameters of the optical cable material and environmental constraints to construct a finite element analysis model;

[0116] For each routing scheme in the population, its performance in the actual environment is simulated through the physics engine, multiple indicators are output and the fitness value is calculated by weighting, and the routing schemes whose fitness values ​​meet the requirements are retained.

[0117] For the retained routing schemes, two groups are randomly selected as parents, and gene fragments are exchanged using a segmented crossover method. The gene parameters of the offspring are randomly modified with a preset probability to simulate gene mutations in biological evolution. The offspring routing schemes generated by crossover and mutation are verified, and the process is iterated based on the verification results until a new routing form is obtained.

[0118] Please continue to refer to Figure 4 In this embodiment of the application, the method further includes a step of detecting the hidden segment route, specifically including:

[0119] S401: Based on the digital twin model, a routing feature transfer learning module is introduced. The features of non-hidden segments are used as basic samples and transferred to the hidden segment sub-model of the digital twin model through federated learning. Combined with the indirect feature data around the hidden segments collected by the dynamic layer, an implicit mapping model of the environmental routing state is trained and generated.

[0120] In this embodiment, the environmental features and routing state correlation patterns of the non-hidden segment are used as basic samples, and these patterns are transferred to the hidden segment sub-model through transfer learning to solve the problem of insufficient hidden segment samples.

[0121] In some embodiments, if the data around the hidden section is distributed in different operation and maintenance units, federated learning is used to achieve data not being shared but model co-training. Each unit trains its sub-model locally and only uploads the model parameters to the central node for aggregation, which protects data privacy and can integrate indirect data from multiple sources.

[0122] S402: Based on the digital twin model, a multimodal data coupling network is constructed. The feature vectors of vibration, light signal and geological structure are weighted through the attention mechanism to generate a dynamically updated three-dimensional routing model of the hidden segment.

[0123] Specifically, in step S402, in the multimodal data coupling network, different types of data are converted into feature vectors of a unified dimension (such as vibration feature vectors, optical signal feature vectors, and geological feature vectors); an attention mechanism is introduced to weight the feature vectors: weights are dynamically assigned according to the correlation strength between features and the state of the hidden segment, strengthening the influence of key information; based on the weighted multimodal features, a three-dimensional routing model of the hidden segment with the required proportion is generated in the digital twin model, and the spatial morphology, state identification, and dynamic correlation are updated in real time.

[0124] S403: Based on the three-dimensional routing model of the concealed section, the state evolution of the concealed section under environmental stress is simulated through a digital twin physical engine to mark high-risk areas; among them, high-risk areas are determined by preset thresholds based on the simulation results.

[0125] Furthermore, in step S403, the marking results are synchronized to the interaction layer.

[0126] Preferably, in the digital twin model of this application embodiment, the step of introducing a biomimetic evolution mechanism to form a new routing model morphology adapted to the current environment includes: for concealed routing segments, using known endpoint parameters and ground-penetrating radar point cloud data to establish an inverse mathematical model; by minimizing the objective function of routing curvature and material stress, the most likely continuous routing trajectory is deduced and incorporated into the candidate morphology set.

[0127] Preferably, in the dynamic layer of the digital twin model:

[0128] Dynamic data and basic layer environmental features are mapped to entity nodes in a knowledge graph, and semantic relationships between entities are extracted using sequence labeling methods.

[0129] A hierarchical structure is constructed that includes four types of ontology: physical entities, environmental elements, event types, and indicator parameters. The logical relationships between ontology are defined, and multi-source heterogeneous data are fused using the DS evidence theory.

[0130] The design incorporates a three-tiered rule system comprising domain rules, association rules, and inference rules, with each rule configured with a weight coefficient and a trigger threshold. A forward chain inference algorithm is employed to match real-time data with entities and relationships in the knowledge graph, transforming raw data into interpretable event tags and constructing a three-tiered mapping of data, features, and events.

[0131] A graph neural network that constructs a spatiotemporal attention mechanism integrates the temporal and spatial dimensions into associative reasoning, represented as follows:

[0132]

[0133] In the formula, Let v represent the feature vector of node v at time t. Represents the set of adjacent nodes. Represents the spatiotemporal attention weights. , Represents the learnable weight matrix; This represents the activation function. This represents the eigenvector of node u at time t;

[0134] In this embodiment, the PC algorithm combined with Granger causality test is used to mine the causal relationship between dynamic data and environmental features. In the PC algorithm, an undirected graph is first constructed, and edges that are independent given other variables are removed using an independence test, simplifying the graph structure. The causal direction is determined based on the V-structure (e.g., variables A and C are independent, but related given B, then B is a common cause of A and C), ultimately generating a directed acyclic graph (DAG) to represent the causal relationship between variables. In this embodiment, the PC algorithm is used to eliminate false associations (e.g., simultaneous temperature rise and optical cable attenuation without causation) from dynamic data and environmental features, and to identify true causal relationships (e.g., construction vibration causing slight bending of the optical cable, leading to signal attenuation). In the Granger causality test of this embodiment, for time-series data (e.g., OTDR optical attenuation curves, real-time monitoring data from sensors), the temporal causal relationship between environmental changes (e.g., sudden temperature rise) and optical cable abnormalities (e.g., signal attenuation) is determined, providing temporal causal evidence for the dynamic layer's association reasoning.

[0135] Furthermore, anomaly detection is performed using the Isolation Forest algorithm combined with knowledge graph constraints. When an anomaly is detected, the knowledge graph is queried through the rule engine to extract related environmental features. A formula for calculating the correlation degree of environmental anomalies is designed, expressed as: Correlation degree = λ1S KG +λ2S ST +λ3S C Among them, S KG For the semantic similarity of knowledge graphs, S ST S represents the spatiotemporal co-occurrence frequency. C λ1, λ2, and λ3 represent the causal strength, and λ3 are weight coefficients that are dynamically adjusted through reinforcement learning. Based on the correlation evaluation results, the sensitivity of the model to different data is dynamically adjusted.

[0136] In some embodiments of this application, in the step of extracting semantic relationships between entities using sequence labeling methods, a sequence labeling model that integrates dynamic attention mechanisms and knowledge graph constraints is used to extract entity semantic relationships. The sequence labeling model is represented as follows:

[0137]

[0138] In the formula, x represents the input text sequence. Used to capture contextual information at the i-th position of the sequence. This represents the attention weight at position i calculated through the attention mechanism, used to dynamically adjust the weights at different positions. CRF stands for Conditional Random Field Function, used to handle dependencies between labels.

[0139] Furthermore, embodiments of this application introduce a loss function L based on knowledge graph constraints. KG It is represented by cosine similarity, where the loss function L KG Represented as:

[0140]

[0141] In the formula, For the predicted relation vector, This represents the corresponding relation vector in the knowledge graph.

[0142] like Figure 5 As shown, according to another embodiment of this application, a communication optical cable routing detection system based on digital twins is provided. The communication optical cable routing detection system provided in this embodiment includes the following modules:

[0143] Information integration module 501 is used to acquire the marker location information in the communication optical cable and integrate multi-source information based on the differentiated spatiotemporal calibration algorithm;

[0144] The twin model construction module 502 is used to construct a digital twin model comprising a base layer, a dynamic layer, and an interaction layer. In the base layer, a multi-dimensional base model with dual mappings is constructed. Based on the mapping of optical cable physical parameters to geographic coordinates, the inherent characteristics of the environment along the route are incorporated to form a composite base framework including spatial topological relationships, environmental constraint boundaries, and physical characteristic benchmarks. In the dynamic layer, after real-time access to multi-source dynamic data, a knowledge graph and rule engine are introduced to perform semantic parsing of the dynamic data. Simultaneously, a correlation reasoning model between dynamic data and environmental features in the base layer is constructed. When data anomalies are detected, the potential correlation between the anomalies and environmental features is mined, and the model's sensitivity to different data is dynamically adjusted based on the reasoning results. In the interaction layer, an operation and maintenance scenario feature library is constructed, and scenario adaptation and pre-feedback mechanisms are introduced.

[0145] The route detection module 503 is used to overlay a virtual model of the optical cable route onto the physical scene based on the visualization elements generated by the interaction layer, perform detection operations on the target route segment, and feed the detection data back to the dynamic layer of the digital twin model in real time to update the status of the virtual model. The dynamic layer of the digital twin model is used to analyze multi-source dynamic data, and combined with the composite basic framework of the basic layer, a route status evaluation system is constructed. Among them, an abnormal status of the route is identified through a preset threshold judgment mechanism.

[0146] The specific functions and examples of each module and submodule of the apparatus in this disclosure can be found in the relevant descriptions of the corresponding steps in the above method embodiments, and will not be repeated here.

[0147] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers.

[0148] Electronic devices can also refer to various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0149] Electronic devices include a computing unit, which can perform various appropriate actions and processes based on a computer program stored in read-only memory or loaded from a storage unit into random access memory. RAM can also store various programs and data required for device operation. The computing unit, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0150] Multiple components in the device are connected to the I / O interface, including: input units, such as keyboards and mice; output units, such as various types of displays and speakers; storage units, such as disks and optical discs; and communication units, such as network cards, modems, and wireless transceivers.

[0151] The communication unit allows the device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0152] The computing unit can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities.

[0153] Examples of computing units include, but are not limited to, central processing units (CPUs), graphics processing units (GPUs), various dedicated artificial intelligence (AI) computing chips, various computing units that run machine learning model algorithms, digital signal processors (DSPs), and any suitable processors, controllers, microcontrollers, etc.

[0154] The computing unit executes the various methods and processes described above, such as a digital twin-based method for detecting optical fiber routing. For example, in some embodiments, a digital twin-based method for detecting optical fiber routing may be implemented as a computer software program tangibly contained in a machine-readable medium, such as a storage unit.

[0155] In some embodiments, part or all of the computer program may be loaded into and / or installed on the device via ROM and / or a communication unit. When the computer program is loaded into RAM and executed by the computing unit, one or more steps of the digital twin-based optical fiber routing detection method described above can be performed.

[0156] Alternatively, in other embodiments, the computing unit may be configured, by any other suitable means (e.g., by means of firmware), to perform a digital twin-based optical fiber routing detection method.

[0157] Various implementations of the systems and techniques described above in this document can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays, application-specific integrated circuits, application-specific standard products, systems-on-a-chip systems, payload programmable logic devices, computer hardware, firmware, software, and / or combinations thereof.

[0158] These various implementations may include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, and can receive data and instructions from a storage system, at least one input device and at least one output device, and transmit data and instructions to the storage system, the at least one input device and the at least one output device.

[0159] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0160] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory, optical fibers, portable compact disk read-only memory, optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0161] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0162] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0163] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0164] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0165] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for communication cable route detection based on digital twinning, characterized in that, Includes the following steps: Obtain the location information of markers in communication optical cables and integrate multi-source information based on a differentiated spatiotemporal calibration algorithm; A digital twin model is constructed, comprising a base layer, a dynamic layer, and an interaction layer. In the base layer, a multi-dimensional base model with dual mappings is built. Based on the mapping of optical cable physical parameters and geographic coordinates, the inherent characteristics of the environment along the route are incorporated to form a composite base framework that includes spatial topological relationships, environmental constraint boundaries, and physical characteristic benchmarks. In the dynamic layer, after real-time access to multi-source dynamic data, a knowledge graph and rule engine are introduced to perform semantic parsing of the dynamic data. At the same time, an association reasoning model between dynamic data and environmental features of the base layer is constructed. When data anomalies are detected, the potential association between anomalies and environmental features is explored, and the sensitivity of the model to different data is dynamically adjusted based on the reasoning results. In the interaction layer, an operation and maintenance scenario feature library is constructed, and a scenario adaptation and pre-feedback mechanism is introduced, including: decomposing the operation and maintenance scenario into triplet feature vectors, establishing a mapping relationship between the scenario and multi-source data through a knowledge graph; dynamically adjusting the weights of each parameter in the feature library based on historical operation and maintenance records using a reinforcement learning algorithm; introducing a scenario lifecycle to construct a scenario evolution state machine, with each state corresponding to interaction rules and data priority configurations; collecting on-site environmental information through sensors, combining it with real-time data from the digital twin model, using multimodal fusion to identify the current operation and maintenance scenario, retrieving corresponding parameter configurations from the feature library based on the identified scenario, and generating the optimal set of interaction rules through a genetic algorithm; dynamically reconstructing the interactive interface using a responsive design algorithm based on the hardware parameters of the interactive device and the current scenario requirements; responding to the detection operation performed by the user, combining the current scenario features, using the physical engine of the digital twin model to simulate the results of the detection operation, and establishing a pre-feedback decision tree to conduct a risk assessment of the simulation results; comparing the actual user operation results with the pre-feedback predictions, and optimizing the pre-feedback model parameters through a meta-learning algorithm; Based on the visualization elements generated by the interaction layer, the virtual model of the optical cable route is overlaid on the physical scene to perform detection operations on the target route segment. The detection data is fed back to the dynamic layer of the digital twin model in real time to update the status of the virtual model. The dynamic layer of the digital twin model is used to analyze multi-source dynamic data. Combined with the composite basic framework of the basic layer, a route status evaluation system is constructed, and abnormal status of the route is identified through a preset threshold judgment mechanism.

2. The digital-twin-based communication cable route detection method of claim 1, wherein, Multi-source information includes OTDR light pulse reflection data, GIS geographic coordinate data, fiber optic distributed sensing data, and manual inspection records; The step of integrating multi-source information based on the differential spatiotemporal calibration algorithm includes spatial calibration and temporal calibration. The spatially calibrated multi-source data is superimposed and displayed in a three-dimensional coordinate system. The calibration effect is verified by calculating the spatial distance deviation between data points and the correlation coefficient of the time series until the deviation does not exceed the threshold. in: In spatial calibration, for OTDR optical pulse reflection data, an optical path-to-actual distance conversion model is established based on the physical parameters of the optical cable, converting the optical path data calculated by the optical pulse transmission time into the actual geographic distance. Combining the topological relationships of the GIS vector map, coordinate mapping is used to anchor the OTDR distance data to the GIS two-dimensional coordinate system, achieving spatial alignment. For fiber optic grating distributed sensing data, the GPS coordinates of the sensor deployment locations are extracted and matched with the optical cable routing nodes in the GIS map to correct spatial offsets caused by sensor installation errors, ensuring a one-to-one correspondence between the sensing data and the optical cable routing locations. Unstructured location descriptions in manual inspection records are parsed into geographic coordinates using natural language processing technology and compared with GIS data for calibration. In time calibration, the Dynamic Time Warping (DTW) algorithm is used to stretch or compress the time axis of time-series data to address the differences in sampling frequencies of different devices; the time information recorded by manual inspection is aligned with the time axis through time standardization.

3. The communication optical cable route detection method based on digital twin according to claim 2, characterized in that, In digital twin models, the steps involved in introducing biological mimicry evolutionary mechanisms to form new routing model forms adapted to the current environment include: Data were collected on typical routing characteristics of optical cables under different environments. Principal component analysis was used to map environmental parameters and routing characteristics into multidimensional feature vectors to construct a routing gene library for communication optical cables. When the digital twin model receives new detection data or environmental information, it retrieves the corresponding feature parameter set as the initial solution. A genetic algorithm is used to perform crossover and mutation operations on the retrieved feature parameters. Combined with the mechanical properties of optical cable materials, physical engine constraints are constructed to simulate the stress and deformation process of optical cable under new environment and generate multiple sets of candidate route forms. By optimizing multiple objectives such as minimizing transmission loss, maximizing mechanical stability, and minimizing maintenance costs, the optimal solution is selected from the candidate model set to form a new routing model that adapts to the current environment.

4. The communication optical cable route detection method based on digital twin according to claim 3, characterized in that, The steps for generating multiple candidate route patterns include: The optical cable routing pattern is decomposed into quantifiable genetic parameters, and a genetic chain is formed by two-dimensional encoding of parameters and thresholds. Historical routing parameters with a matching degree greater than a threshold for the current environment characteristics are retrieved from the routing gene library, and several initial routing schemes are generated by random combination. Import the mechanical parameters of the optical cable material and environmental constraints to construct a finite element analysis model; For each routing scheme in the population, its performance in the actual environment is simulated through the physics engine, multiple indicators are output and the fitness value is calculated by weighting, and the routing schemes whose fitness values ​​meet the requirements are retained. For the retained routing schemes, two groups are randomly selected as parents, and gene fragments are exchanged in a segmented crossover manner. The gene parameters of the offspring are randomly modified with a preset probability to simulate gene mutations in biological evolution. The offspring routing schemes generated by crossover and mutation are verified, and the process is iterated based on the verification results until a new routing form is obtained.

5. The communication optical cable route detection method based on digital twin according to claim 4, characterized in that, It also includes the step of detecting hidden segment routes, including: Based on the digital twin model, a routing feature transfer learning module is introduced. The features of non-hidden segments are used as basic samples and transferred to the hidden segment sub-model of the digital twin model through federated learning. Combined with the indirect feature data around the hidden segments collected by the dynamic layer, an implicit mapping model of the environmental routing state is trained and generated. A multimodal data coupling network is constructed based on a digital twin model. The feature vectors of vibration, optical signals and geological structures are weighted through an attention mechanism to generate a dynamically updated three-dimensional routing model for hidden segments. Based on the three-dimensional routing model of the concealed segment, the state evolution of the concealed segment under environmental stress is simulated through a digital twin physical engine, and high-risk areas are marked.

6. The communication optical cable route detection method based on digital twin according to claim 5, characterized in that, In digital twin models, the steps of introducing biological mimicry evolutionary mechanisms to form new routing model forms adapted to the current environment also include: For concealed route segments, an inverse mathematical model is established using known endpoint parameters and ground-penetrating radar point cloud data; By minimizing the objective function of routing curvature and material stress, the most likely continuous routing trajectory is derived and incorporated into the candidate morphology set.

7. The communication optical cable route detection method based on digital twin according to claim 6, characterized in that, In the dynamic layer of the digital twin model: Dynamic data and basic layer environmental features are mapped to entity nodes in a knowledge graph. Semantic relationships between entities are extracted using sequence labeling. A hierarchical structure containing four types of ontology—physical entities, environmental elements, event types, and indicator parameters—is constructed. Logical relationships between ontology are defined. Multi-source heterogeneous data are fused using DS evidence theory. A three-layer rule system is constructed, comprising domain rules, association rules, and inference rules. Each rule is configured with a weight coefficient and a trigger threshold. A forward chain inference algorithm is used to match real-time data with entities and relationships in the knowledge graph, converting raw data into interpretable event tags and constructing a three-layer mapping of data-feature-event. A graph neural network that constructs a spatiotemporal attention mechanism integrates the temporal and spatial dimensions into associative reasoning, represented as follows: ； In the formula, Let v represent the feature vector of node v at time t. Represents the set of adjacent nodes. Represents the spatiotemporal attention weights. , Represents the learnable weight matrix; This represents the activation function. This represents the eigenvector of node u at time t; We employ PC algorithm combined with Granger causality test to explore the causal relationship between dynamic data and environmental features; Anomaly detection is performed using the Isolation Forest algorithm combined with knowledge graph constraints. When an anomaly is detected, the knowledge graph is queried through a rule engine to extract related environmental features. The formula for calculating the environmental anomaly correlation is: Correlation Degree. ， in, For semantic similarity of knowledge graphs, Indicates the spatiotemporal co-occurrence frequency. For causal strength, These are the weighting coefficients, which are dynamically adjusted through reinforcement learning; based on the correlation evaluation results, the model's sensitivity to different data is dynamically adjusted.

8. The communication optical cable route detection method based on digital twin according to claim 7, characterized in that, In the step of extracting semantic relationships between entities using sequence labeling, a sequence labeling model that integrates dynamic attention mechanism and knowledge graph constraints is employed to extract entity semantic relationships. in The sequence labeling model is represented as: ; In the formula, x represents the input text sequence. Used to capture contextual information at the i-th position of the sequence. This represents the attention weight at position i, calculated using the attention mechanism, which is used to dynamically adjust the weights at different positions. This represents a conditional random field function used to handle dependencies between labels; Loss function with knowledge graph constraints It is represented by cosine similarity, where the loss function is... Represented as: ; In the formula, For the predicted relation vector, This represents the corresponding relation vector in the knowledge graph.

9. A detection system for implementing the communication optical cable route detection method based on digital twins as described in any one of claims 1 to 8, characterized in that, The detection system includes the following modules: The information integration module is used to acquire the marker location information in the communication optical cable and integrate multi-source information based on the differentiated spatiotemporal calibration algorithm; The twin model construction module is used to build a digital twin model comprising a base layer, a dynamic layer, and an interaction layer. In the base layer, a multi-dimensional base model with dual mappings is constructed. Based on the mapping of optical cable physical parameters to geographic coordinates, the inherent characteristics of the environment along the route are incorporated to form a composite base framework including spatial topological relationships, environmental constraint boundaries, and physical characteristic benchmarks. In the dynamic layer, after real-time access to multi-source dynamic data, a knowledge graph and rule engine are introduced to perform semantic parsing of the dynamic data. Simultaneously, a correlation reasoning model between dynamic data and environmental features in the base layer is constructed. When data anomalies are detected, the potential correlation between the anomalies and environmental features is mined, and the model's sensitivity to different data is dynamically adjusted based on the reasoning results. In the interaction layer, an operation and maintenance scenario feature library is constructed, and scenario adaptation and pre-feedback mechanisms are introduced. The route detection module is used to overlay a virtual model of the optical cable route onto the physical scene based on the visualization elements generated by the interaction layer, perform detection operations on the target route segment, and feed the detection data back to the dynamic layer of the digital twin model in real time to update the status of the virtual model. The dynamic layer of the digital twin model is used to analyze multi-source dynamic data, and combined with the composite basic framework of the basic layer, a route status evaluation system is constructed. An abnormal state of the route is identified through a preset threshold judgment mechanism.

Images