Online monitoring method and system for hollow core anti-resonance optical fiber communication system

Abstract

The present invention discloses an online monitoring method and system for a hollow-core anti-resonant optical fiber communication system, and relates to the field of optical fiber communication monitoring technology. After extracting the spatiotemporal features of optical fiber transmission data and environmental data using a spatiotemporal analysis model, the features of the optical fiber signal in the time and space dimensions are constructed into a spatiotemporal matrix, and the optical fiber communication system is modeled as a graph structure. The spatiotemporal matrix and the graph structure are input into a pre-trained spatiotemporal graph convolutional network. The spatiotemporal graph convolutional network captures the dependencies between nodes and the spatiotemporal dynamic changes in the graph structure, and predicts whether there are abnormalities in the optical fiber communication system. If there are abnormalities, the monitoring system automatically executes corresponding response measures and triggers an alarm. The monitoring system extracts the spatiotemporal features of the optical fiber transmission data using a spatiotemporal analysis model, and captures the dependencies between nodes and the spatiotemporal dynamic changes in the graph structure through a spatiotemporal graph convolutional network, thereby improving the sensitivity to abnormal situations and the accuracy of detection.

Description

Technical Field

[0001] The present invention relates to the field of optical fiber communication monitoring technology, and in particular to an online monitoring method and system for a hollow core anti-resonance optical fiber communication system. Background Art

[0002] Hollow-Core-Anti-Resonant0Fiber (HC-ARF) is a new type of optical fiber that achieves low-loss, high-efficiency light transmission by guiding light waves in the fiber core. Unlike traditional solid-core optical fibers, the light waves of hollow-core anti-resonant optical fibers mainly propagate in the hollow fiber core, relying on the anti-resonance effect of the fiber wall to suppress light leakage, thereby achieving the purpose of efficient transmission. This type of optical fiber has broad application prospects in high-speed data communication, laser transmission, sensing and other fields. In hollow-core anti-resonant optical fiber communication systems, real-time online monitoring is required to ensure the stability and reliability of the communication link.

[0003] The prior art has the following deficiencies:

[0004] 1. Traditional fiber optic monitoring methods often rely on feature analysis in a single time or space dimension and cannot effectively capture the dynamic changes of signals in time and space dimensions. This results in poor detection accuracy and timeliness when facing complex transmission environments and signal fluctuations.

[0005] 2. The monitoring system only focuses on the transmission characteristics of the optical fiber signal, but ignores the impact of environmental factors on the optical fiber performance, making it difficult to identify potential faults caused by environmental changes. Summary of the invention

[0006] The purpose of the present invention is to provide an online monitoring method and system for a hollow-core anti-resonant optical fiber communication system, which utilizes a space-time analysis model to extract space-time features of optical fiber transmission data, and captures the dependencies between nodes and the space-time dynamic changes in the graph structure through a space-time graph convolutional network, thereby improving the sensitivity to abnormal situations and the accuracy of detection. Environmental data is incorporated into node features in the graph structure and analyzed in combination with the transmission characteristics of optical fiber signals, so that the system can comprehensively consider the impact of the environment on optical fiber performance, thereby improving the accuracy of fault prediction and addressing the shortcomings of the background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: an online monitoring method for a hollow core anti-resonant optical fiber communication system, the monitoring method comprising the following steps:

[0008] The monitoring system collects optical fiber transmission data and environmental data associated with the optical fiber communication system in real time;

[0009] After extracting the spatiotemporal features of the optical fiber transmission data and environmental data using the spatiotemporal analysis model, the characteristics of the optical fiber signal in the time and space dimensions are constructed into a spatiotemporal matrix, and the optical fiber communication system is modeled as a graph structure.

[0010] The space-time matrix and graph structure are input into the pre-trained space-time graph convolutional network. The space-time graph convolutional network captures the dependencies between nodes and the spatiotemporal dynamic changes in the graph structure, and predicts whether there are any abnormalities in the optical fiber communication system. If there are any abnormalities, the monitoring system automatically executes corresponding response measures and triggers alarms, and marks the abnormal areas or nodes in the graph structure and displays them to relevant personnel.

[0011] In a preferred embodiment, the monitoring system collects in real time the optical fiber transmission data and environmental data associated with the optical fiber communication system, the optical fiber transmission data including spectral width, polarization mode dispersion, and optical signal intensity floating assignment, and the environmental data including temperature deviation and stress accumulation coefficient.

[0012] In a preferred embodiment, after extracting the spatiotemporal features of the optical fiber transmission data and the environmental data using the spatiotemporal analysis model, the features of the optical fiber signal in the time and space dimensions are constructed into a spatiotemporal matrix, including the following steps:

[0013] After normalizing the optical fiber transmission data and environmental data, the optical fiber signal index is obtained by comprehensive calculation, and the expression is:

[0014] , where is the fiber signal index, For optical fiber data transmission, For environmental data, , are the adjustment coefficients of optical fiber transmission data and environmental data, respectively, and , are greater than 0, is the spectrum width, is the polarization mode dispersion, Assign a floating value to the optical signal strength, is the temperature deviation, is the stress accumulation coefficient;

[0015] The fiber signal index of nodes at different time points in the hollow-core anti-resonant fiber communication system is obtained as the time feature, and the fiber signal index of each node at the same time point is obtained as the space feature, and the time feature and space feature are combined to establish a time-space feature matrix.

[0016] In a preferred embodiment, the spatiotemporal matrix and the graph structure are input into a pre-trained spatiotemporal graph convolutional network to predict whether there is an abnormality in the optical fiber communication system, including the following steps:

[0017] The spatiotemporal matrix and graph structure are used as the input of the spatiotemporal graph convolutional network. The spatiotemporal matrix provides temporal and spatial feature information, and the graph structure provides the connection relationship and interdependence between nodes.

[0018] The spatial convolution is performed on the graph structure through the graph convolution operation to capture the spatial dependencies between nodes. The graph convolution propagates the node features through the adjacency matrix, so that the features of each node are fused with the features of its neighboring nodes to extract local spatial features.

[0019] Perform a temporal convolution operation on the spatiotemporal matrix to capture the changes in features in the time dimension. Temporal convolution extracts the time dependency information from time series data through a one-dimensional convolution method.

[0020] After multiple layers of spatiotemporal convolution, the spatiotemporal features extracted from each layer are aggregated to obtain the aggregated feature index, which is expressed as: , where is the aggregation feature index, is the number of layers of the spatiotemporal graph convolutional network, For the Layer extracted fiber signal index;

[0021] The obtained aggregate feature index is input into the fully connected layer, and the fully connected layer compares the obtained aggregate feature index with a preset index threshold, and the index threshold is used to predict whether the hollow core anti-resonance optical fiber communication system will be abnormal;

[0022] If the aggregate characteristic index is greater than or equal to the index threshold, it is predicted that the hollow-core anti-resonance optical fiber communication system will not be abnormal. If the aggregate characteristic index is less than the index threshold, it is predicted that the hollow-core anti-resonance optical fiber communication system will be abnormal.

[0023] In a preferred embodiment, the optical fiber communication system is modeled as a graph structure, comprising the following steps:

[0024] Nodes represent key points in optical fiber signals, including signal collection points, optical amplifiers, or optical fiber connection points. Edges represent the transmission paths of optical fiber signals. Each edge connects two nodes, indicating the path of signal transmission from one key point to another.

[0025] Assign initial characteristic values ​​to nodes. Each node should include its related optical fiber transmission data, including spectrum width, polarization mode dispersion, and optical signal strength floating assignment. Each node also includes related environmental data, including temperature deviation and stress accumulation coefficient.

[0026] Each edge includes loss characteristics associated with optical signal transmission, as well as the physical distance between nodes.

[0027] In a preferred embodiment, the calculation expression of the spectrum width is: , where is the spectrum width, is the wavelength of the optical signal, is the central wavelength of the spectrum, usually indicating the center of gravity or peak wavelength of the spectrum. The wavelength of the optical signal Spectral power density distribution at ;

[0028] The calculation expression of the polarization mode dispersion is: , where is the polarization mode dispersion, is the group delay of an optical signal propagating along one polarization direction in an optical fiber, is the group delay of the optical signal propagating along another orthogonal polarization direction in the optical fiber, is the average of the squares of the group delay differences in two orthogonal polarization directions;

[0029] The calculation logic of the optical signal intensity floating assignment is: obtain the optical signal intensity through multiple monitoring points of the hollow-core anti-resonance optical fiber communication system, calculate the average optical signal intensity and the intensity standard deviation based on the optical signal intensity obtained from the multiple monitoring points, and obtain the optical signal intensity floating assignment by dividing the average optical signal intensity by the intensity standard.

[0030] In a preferred embodiment, the calculation logic of the temperature deviation is: mark the temperature stability range as , and mark the real-time monitored temperature as ,when When the temperature deviation ;when When the temperature deviation ,when When the temperature deviation ;

[0031] The calculation expression of the stress accumulation coefficient is: , where represents the stress accumulation factor, Indicates at time The stress of the moment, is the start time of accumulation, is the cumulative end time.

[0032] The online monitoring system of hollow core anti-resonant optical fiber communication system includes a space-time matrix building module, a graph structure modeling module, and an abnormality prediction module:

[0033] Space-time matrix construction module: collects optical fiber transmission data and environmental data associated with the optical fiber communication system in real time, extracts the space-time features of the optical fiber transmission data and environmental data using the space-time analysis model, and constructs the features of the optical fiber signal in the time and space dimensions into a space-time matrix;

[0034] Graph structure modeling module: The optical fiber communication system is modeled as a graph structure. The nodes in the graph structure represent the key points in the optical fiber signal, and the edges represent the transmission paths of the optical fiber signal. Initial feature values ​​are assigned to each node and edge in the graph structure. The node features include local optical signal characteristics and environmental data, and the edge features include transmission loss and distance.

[0035] Abnormality prediction module: The space-time matrix and graph structure are input into the pre-trained space-time graph convolutional network. The space-time graph convolutional network captures the dependencies between nodes and the dynamic changes in space and time in the graph structure, and predicts whether there are abnormalities in the optical fiber communication system. If there are abnormalities, the corresponding response measures are automatically executed and the alarm is triggered. The abnormal areas or nodes are marked in the graph structure and displayed to relevant personnel.

[0036] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0037] After the present invention uses the spatiotemporal analysis model to extract the spatiotemporal features of the optical fiber transmission data and the environmental data, the features of the optical fiber signal in the time and space dimensions are constructed into a spatiotemporal matrix, the optical fiber communication system is modeled as a graph structure, and the spatiotemporal matrix and the graph structure are input into a pre-trained spatiotemporal graph convolutional network. The spatiotemporal graph convolutional network captures the dependencies between nodes and the spatiotemporal dynamic changes in the graph structure, and predicts whether there are abnormalities in the optical fiber communication system. If there are abnormalities, the monitoring system automatically executes corresponding response measures and triggers an alarm. The monitoring system uses the spatiotemporal analysis model to extract the spatiotemporal features of the optical fiber transmission data, and captures the dependencies between nodes and the spatiotemporal dynamic changes in the graph structure through the spatiotemporal graph convolutional network, which improves the sensitivity to abnormal situations and the accuracy of detection. The environmental data is incorporated into the node features in the graph structure, and is analyzed in combination with the transmission characteristics of the optical fiber signal, so that the system can fully consider the impact of the environment on the optical fiber performance, and improve the accuracy of fault prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0039] Figure 1 The present invention is a flow chart of the method. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] Example 1: Please refer to Figure 1 As shown, the online monitoring method of the hollow core anti-resonant optical fiber communication system described in this embodiment includes the following steps:

[0042] The monitoring system collects fiber transmission data and environmental data associated with the fiber optic communication system in real time. After extracting the spatiotemporal features of the fiber transmission data and environmental data using the spatiotemporal analysis model, the features of the fiber optic signal in the time and space dimensions are constructed into a spatiotemporal matrix, and the fiber optic communication system is modeled as a graph structure, in which nodes represent key points in the fiber optic signal (such as signal acquisition points, amplifiers, etc.), and edges represent the transmission path of the fiber optic signal. An initial feature value is assigned to each node and edge in the graph structure. Node features may include local optical signal characteristics and environmental data, and edge features may include transmission loss, distance, etc. The spatiotemporal matrix and the graph structure are input into a pre-trained spatiotemporal graph convolutional network. The spatiotemporal graph convolutional network captures the dependencies between nodes and the spatiotemporal dynamic changes in the graph structure, and predicts whether there are abnormalities in the fiber optic communication system. If there are abnormalities, the monitoring system automatically executes corresponding response measures and triggers alarms, and displays them to relevant personnel after marking abnormal areas or nodes in the graph structure.

[0043] This application uses a spatiotemporal analysis model to extract spatiotemporal features from optical fiber transmission data and environmental data, constructs the features of optical fiber signals in time and space dimensions into a spatiotemporal matrix, models the optical fiber communication system as a graph structure, and inputs the spatiotemporal matrix and the graph structure into a pre-trained spatiotemporal graph convolutional network. The spatiotemporal graph convolutional network captures the dependencies between nodes and the spatiotemporal dynamic changes in the graph structure, and predicts whether there are abnormalities in the optical fiber communication system. If there are abnormalities, the monitoring system automatically executes corresponding response measures and triggers alarms. The monitoring system uses a spatiotemporal analysis model to extract spatiotemporal features from optical fiber transmission data, and uses a spatiotemporal graph convolutional network to capture the dependencies between nodes and the spatiotemporal dynamic changes in the graph structure, thereby improving the sensitivity to abnormal situations and the accuracy of detection. Environmental data is incorporated into node features in the graph structure, and analyzed in combination with the transmission characteristics of optical fiber signals, so that the system can fully consider the impact of the environment on optical fiber performance and improve the accuracy of fault prediction.

[0044] Embodiment 2: The monitoring system collects optical fiber transmission data and environmental data associated with the optical fiber communication system in real time, the optical fiber transmission data includes spectrum width, polarization mode dispersion, and optical signal intensity floating assignment, and the environmental data includes temperature deviation and stress accumulation coefficient;

[0045] The calculation expression of the spectrum width is: , where is the spectrum width, is the wavelength of the optical signal, is the central wavelength of the spectrum, usually indicating the center of gravity or peak wavelength of the spectrum. The wavelength of the optical signal The spectral power density distribution at , that is, the spectral intensity function, is usually a function that varies with wavelength, indicating the power distribution at different wavelengths;

[0046] Spectral width usually indicates the spread of the optical signal in the frequency domain. A larger spectral width may indicate the presence of dispersion effects or modal spurious in the system, especially in hollow-core antiresonant fibers, where dispersion effects may cause signal distortion and increased transmission loss.

[0047] The calculation expression of polarization mode dispersion is: , where is the polarization mode dispersion, is the group delay of an optical signal propagating along one polarization direction in an optical fiber, is the group delay of the optical signal propagating along another orthogonal polarization direction in the optical fiber, is the average of the squares of the group delay differences in two orthogonal polarization directions;

[0048] Polarization mode dispersion describes the time delay difference between different polarization modes of optical signals in optical fibers. Large polarization mode dispersion indicates that the time delay difference between different polarization states of the signal during transmission is large, which may cause signal distortion and increase bit error rate, especially at high data rate transmission.

[0049] The calculation logic of the optical signal intensity floating assignment is: obtain the optical signal intensity at multiple monitoring points of the hollow-core anti-resonance optical fiber communication system, and calculate the average optical signal intensity and the intensity standard deviation based on the optical signal intensity obtained at multiple monitoring points. The expression is: , where is the average optical signal intensity, is the strength standard deviation, For the The optical signal strength at each monitoring point is is the number of monitoring points, and the average optical signal intensity is divided by the intensity standard to obtain the optical signal intensity floating assignment. The larger the average optical signal intensity is, the smaller the optical signal intensity floating assignment is, indicating that the overall optical signal intensity of the hollow-core antiresonant optical fiber communication system is stronger and the fluctuation is smaller.

[0050] The calculation logic of temperature deviation is: mark the temperature stability range as , and mark the real-time monitored temperature as ,when When the temperature deviation ;when When the temperature deviation ,when When the temperature deviation ;

[0051] Temperature changes affect the refractive index of the optical fiber and the mechanical properties of the optical fiber material, resulting in changes in the optical fiber transmission performance. The greater the temperature deviation, the more likely the system performance will be affected, resulting in increased transmission loss, wavelength drift, and mechanical damage to the optical fiber.

[0052] The calculation expression of stress accumulation coefficient is: , where represents the stress accumulation factor, Indicates at time The stress of the moment, is the start time of accumulation, is the cumulative end time;

[0053] The stress accumulation coefficient indicates the amount of stress accumulated by the optical fiber within a certain time range. A higher stress accumulation coefficient indicates that the optical fiber is under greater mechanical stress for a long time, which may cause permanent deformation, damage or breakage of the optical fiber, affecting signal transmission.

[0054] After extracting the spatiotemporal features of the optical fiber transmission data and the environmental data using the spatiotemporal analysis model, the features of the optical fiber signal in the time and space dimensions are constructed into a spatiotemporal matrix, including the following steps:

[0055] After normalizing the optical fiber transmission data and environmental data, the optical fiber signal index is obtained by comprehensive calculation, and the expression is:

[0056] , where is the fiber signal index, For optical fiber data transmission, For environmental data, , are the adjustment coefficients of optical fiber transmission data and environmental data, respectively, and , are greater than 0, is the spectrum width, is the polarization mode dispersion, Assign a floating value to the optical signal strength, is the temperature deviation, is the stress accumulation factor.

[0057] Obtaining the optical fiber signal index of the nodes at different time points in the hollow core anti-resonance optical fiber communication system as a time feature, and obtaining the optical fiber signal index of each node at the same time point as a spatial feature;

[0058] In the hollow core anti-resonant optical fiber communication system, the time characteristics and space characteristics are combined to establish a time-space characteristic matrix.

[0059] Modeling the optical fiber communication system as a graph structure includes the following steps:

[0060] In the fiber-optic communication system, nodes represent key points in the fiber-optic signal, such as signal collection points, optical amplifiers, fiber connection points, etc. These nodes are the basic elements of the graph structure. Each node corresponds to a physical location or function in the fiber-optic transmission process. Edges represent the transmission path of the fiber-optic signal. Each edge connects two nodes and represents the path of the signal from one key point to another. These edges can be unidirectional (directed graph) or bidirectional (undirected graph), depending on the transmission characteristics of the signal.

[0061] Assign initial characteristic values ​​to nodes. Each node should include its related optical fiber transmission data, including spectrum width, polarization mode dispersion, and optical signal strength floating assignment. Each node should also include related environmental data, including temperature deviation and stress accumulation coefficient.

[0062] Each edge should include the loss characteristics associated with the transmission of optical signals, and should also include the physical distance between nodes, which can affect the attenuation and delay of the signal.

[0063] Use Diagram represents a hollow-core anti-resonant optical fiber communication system, where is a collection of nodes, is a set of edges, each node Has a set of characteristics , each edge Has a set of characteristics , construct an adjacency matrix A or adjacency table to represent the connection relationship between nodes, and the elements of the adjacency matrix Indicates whether there is an edge between node i and node j, if yes, it is 1, otherwise it is 0.

[0064] The spatiotemporal matrix and graph structure are input into the pre-trained spatiotemporal graph convolutional network. The spatiotemporal graph convolutional network captures the dependencies between nodes and the spatiotemporal dynamic changes in the graph structure, and predicts whether there are abnormalities in the optical fiber communication system, including the following steps:

[0065] The spatiotemporal matrix and graph structure are used together as the input of the spatiotemporal graph convolutional network (ST-GCN). The spatiotemporal matrix provides temporal and spatial feature information, and the graph structure provides the connection relationship and interdependence between nodes.

[0066] Spatial convolution: Graph convolution is used to perform spatial convolution on the graph structure to capture the spatial dependencies between nodes. Graph convolution propagates node features through the adjacency matrix, so that the features of each node are fused with the features of its neighboring nodes to extract local spatial features.

[0067] Temporal convolution: Perform a temporal convolution operation on the spatiotemporal matrix to capture the changes in features in the time dimension. Temporal convolution can be implemented through one-dimensional convolution or recurrent neural network (RNN) to extract the time dependency information in time series data.

[0068] After multiple layers of spatiotemporal convolution, the spatiotemporal features extracted from each layer are aggregated to obtain the aggregated feature index, which is expressed as: , where is the aggregation feature index, is the number of layers of the spatiotemporal graph convolutional network, For the Layer extracted fiber signal index;

[0069] The obtained aggregation feature index is input into the fully connected layer, and the fully connected layer compares the obtained aggregation feature index with the preset index threshold. The index threshold is used to predict whether the hollow-core anti-resonance optical fiber communication system will have an abnormality. If the aggregation feature index is greater than or equal to the index threshold, it is predicted that the hollow-core anti-resonance optical fiber communication system will not have an abnormality. If the aggregation feature index is less than the index threshold, it is predicted that the hollow-core anti-resonance optical fiber communication system will have an abnormality.

[0070] If there is an abnormality, the monitoring system automatically executes the corresponding response measures and triggers the alarm, and displays it to relevant personnel after marking the abnormal area or node in the graph structure, including the following steps:

[0071] Generates an alert message containing the nature of the anomaly, the scope of impact, and recommended response actions.

[0072] The following is an example of the alert content:

[0073] Abnormal type: low fiber signal index;

[0074] Aggregate feature index: (assumed value);

[0075] The index threshold is 0.5;

[0076] Recommended actions: Check fiber connections and calibrate equipment;

[0077] Send alarm: Use notification system (such as email, SMS, system notification) to send alarm message to relevant personnel.

[0078] Automatically execute predefined response actions, such as adjusting amplifier settings for fiber optic signals, switching to backup channels, or enabling data retransmission mechanisms.

[0079] Automatic response example:

[0080] Adjust Settings: Automatically adjust amplifier gain.

[0081] Record response log: Record the details of the automatic response, including response time, action content and execution result log record example;

[0082] Measures: Adjust the amplifier gain;

[0083] Result: Signal strength returned to normal;

[0084] Update the graph structure:

[0085] Mark abnormal nodes or edges in the graph structure. Nodes may include fiber signal collection points or amplifiers, and edges represent the transmission paths of fiber signals.

[0086] Example of markup:

[0087] Abnormal nodes: mark the optical fiber signal collection point or amplifier in red.

[0088] Abnormal edges: Add warning signs to the edges connecting abnormal nodes;

[0089] Updated graph visualization: In the graphical interface of the monitoring system, abnormal areas or nodes are highlighted to make them more conspicuous;

[0090] Color: Abnormal nodes are marked in red, and normal nodes are marked in green.

[0091] Icon: A warning icon is displayed next to the abnormal node.

[0092] Produce detailed reports containing exception descriptions, response measures, processing results, and graph structure updates.

[0093] Anomaly Overview: The fiber signal index is low, affecting nodes X and Y.

[0094] Response measures: Adjust amplifier gain and switch to alternate channel.

[0095] Result: The signal strength returned to normal and the anomaly was resolved.

[0096] Display on the monitoring interface: Display the updated graph structure and detailed reports on the dashboard of the monitoring system to ensure that relevant personnel can view and handle abnormal information in real time.

[0097] Dashboard: Displays detailed information and processing status of abnormal nodes.

[0098] Notification: Abnormal alarm notification pops up on the interface.

[0099] Embodiment 3: The online monitoring system of the hollow core anti-resonant optical fiber communication system described in this embodiment includes a space-time matrix construction module, a graph structure modeling module, and an abnormality prediction module:

[0100] Space-time matrix construction module: collects optical fiber transmission data and environmental data associated with the optical fiber communication system in real time, extracts the space-time features of the optical fiber transmission data and environmental data using the space-time analysis model, and constructs the features of the optical fiber signal in time and space dimensions into a space-time matrix. The space-time matrix is ​​sent to the anomaly prediction module, and the optical fiber transmission data and environmental data are sent to the graph structure modeling module;

[0101] Graph structure modeling module: The optical fiber communication system is modeled as a graph structure, where nodes represent key points in the optical fiber signal (such as signal collection points, amplifiers, etc.), and edges represent the transmission paths of the optical fiber signals. Initial feature values ​​are assigned to each node and edge in the graph structure. Node features may include local optical signal characteristics and environmental data, and edge features may include transmission loss, distance, etc. The graph structure is sent to the anomaly prediction module;

[0102] Abnormality prediction module: The space-time matrix and graph structure are input into the pre-trained space-time graph convolutional network. The space-time graph convolutional network captures the dependencies between nodes and the dynamic changes in space and time in the graph structure, and predicts whether there are abnormalities in the optical fiber communication system. If there are abnormalities, the corresponding response measures are automatically executed and the alarm is triggered. The abnormal areas or nodes are marked in the graph structure and displayed to relevant personnel.

[0103] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0104] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0105] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0106] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the above-described system, device and unit can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0107] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for online monitoring of a hollow core anti-resonant optical fiber communication system, characterized in that: The monitoring method comprises the following steps: The monitoring system collects optical fiber transmission data and environmental data associated with the optical fiber communication system in real time; After extracting the spatiotemporal features of the optical fiber transmission data and environmental data using the spatiotemporal analysis model, the characteristics of the optical fiber signal in the time and space dimensions are constructed into a spatiotemporal matrix, and the optical fiber communication system is modeled as a graph structure. The spatiotemporal matrix and graph structure are input into the pre-trained spatiotemporal graph convolutional network. The spatiotemporal graph convolutional network captures the dependencies between nodes and the spatiotemporal dynamic changes in the graph structure, and predicts whether there are abnormalities in the optical fiber communication system. If there are abnormalities, the monitoring system automatically executes corresponding response measures and triggers alarms, and marks the abnormal areas or nodes in the graph structure and displays them to relevant personnel. After extracting the spatiotemporal features of the optical fiber transmission data and the environmental data using the spatiotemporal analysis model, the features of the optical fiber signal in the time and space dimensions are constructed into a spatiotemporal matrix, including the following steps: After normalizing the optical fiber transmission data and environmental data, the optical fiber signal index is obtained by comprehensive calculation, and the expression is: , where is the fiber signal index, For optical fiber data transmission, For environmental data, are the adjustment coefficients of optical fiber transmission data and environmental data, respectively, and are all greater than 0, Δλ is the spectrum width, is the polarization mode dispersion, Assign a floating value to the optical signal strength, is the temperature deviation, is the stress accumulation coefficient; The fiber signal index of the node at different time points in the hollow core anti-resonance fiber communication system is obtained as the time feature, the fiber signal index of each node at the same time point is obtained as the space feature, and the time feature and the space feature are used to establish a time-space matrix; The spatiotemporal matrix and graph structure are input into the pre-trained spatiotemporal graph convolutional network to predict whether there is an abnormality in the optical fiber communication system, including the following steps: The spatiotemporal matrix and graph structure are used as the input of the spatiotemporal graph convolutional network. The spatiotemporal matrix provides temporal and spatial feature information, and the graph structure provides the connection relationship and interdependence between nodes. The spatial convolution is performed on the graph structure through the graph convolution operation to capture the spatial dependencies between nodes. The graph convolution propagates the node features through the adjacency matrix, so that the features of each node are fused with the features of its neighboring nodes to extract local spatial features. Perform a temporal convolution operation on the spatiotemporal matrix to capture the changes in features in the time dimension. Temporal convolution extracts the time dependency information in the time series data through a one-dimensional convolution method. After multiple layers of spatiotemporal convolution, the spatiotemporal features extracted from each layer are aggregated to obtain the aggregated feature index, which is expressed as: , where is the aggregation feature index, is the number of layers of the spatiotemporal graph convolutional network, For the Layer extracted fiber signal index; The obtained aggregate feature index is input into the fully connected layer, and the fully connected layer compares the obtained aggregate feature index with a preset index threshold, and the index threshold is used to predict whether the hollow core anti-resonance optical fiber communication system will be abnormal; If the aggregate characteristic index is greater than or equal to the index threshold, it is predicted that the hollow core anti-resonance optical fiber communication system will not be abnormal; if the aggregate characteristic index is less than the index threshold, it is predicted that the hollow core anti-resonance optical fiber communication system will be abnormal; Modeling the optical fiber communication system as a graph structure includes the following steps: Nodes represent key points in optical fiber signals, including signal collection points, optical amplifiers, or optical fiber connection points. Edges represent the transmission paths of optical fiber signals. Each edge connects two nodes, indicating the path of signal transmission from one key point to another. Assign initial characteristic values ​​to nodes. Each node should include its related optical fiber transmission data, including spectrum width, polarization mode dispersion, and optical signal strength floating assignment. Each node also includes related environmental data, including temperature deviation and stress accumulation coefficient. Each edge includes loss characteristics associated with optical signal transmission, as well as the physical distance between nodes.

2. The online monitoring method of hollow core anti-resonance optical fiber communication system according to claim 1, characterized in that: The monitoring system collects optical fiber transmission data and environmental data associated with the optical fiber communication system in real time. The optical fiber transmission data includes spectrum width, polarization mode dispersion, and optical signal intensity floating assignment. The environmental data includes temperature deviation and stress accumulation coefficient.

3. The online monitoring method of hollow core anti-resonance optical fiber communication system according to claim 2, characterized in that: The calculation expression of the spectrum width is: , where is the spectrum width, is the wavelength of the optical signal, is the central wavelength of the spectrum, usually indicating the center of gravity or peak wavelength of the spectrum. The wavelength of the optical signal Spectral power density distribution at ; The calculation expression of the polarization mode dispersion is: , where is the polarization mode dispersion, is the group delay of an optical signal propagating along one polarization direction in an optical fiber, is the group delay of the optical signal propagating along another orthogonal polarization direction in the optical fiber, is the average of the squares of the group delay differences in two orthogonal polarization directions; The calculation logic of the optical signal intensity floating assignment is: obtain the optical signal intensity through multiple monitoring points of the hollow-core anti-resonance optical fiber communication system, calculate the average optical signal intensity and the intensity standard deviation based on the optical signal intensity obtained from the multiple monitoring points, and obtain the optical signal intensity floating assignment by dividing the average optical signal intensity by the intensity standard.

4. The online monitoring method of hollow core anti-resonance optical fiber communication system according to claim 3, characterized in that: The calculation logic of the temperature deviation is: mark the temperature stability range as , and mark the real-time monitored temperature as ,when When the temperature deviation ;when When the temperature deviation ,when , temperature deviation ; The calculation expression of the stress accumulation coefficient is: , where represents the stress accumulation factor, Indicates at time The stress of the moment, is the start time of accumulation, is the cumulative end time.

5. An online monitoring system for a hollow core anti-resonant optical fiber communication system, used to implement the monitoring method according to any one of claims 1 to 4, characterized in that: Including space-time matrix construction module, graph structure modeling module, and anomaly prediction module: Space-time matrix construction module: collects optical fiber transmission data and environmental data associated with the optical fiber communication system in real time, extracts the space-time features of the optical fiber transmission data and environmental data using the space-time analysis model, and constructs the features of the optical fiber signal in the time and space dimensions into a space-time matrix; Graph structure modeling module: The optical fiber communication system is modeled as a graph structure. The nodes in the graph structure represent the key points in the optical fiber signal, and the edges represent the transmission paths of the optical fiber signal. Initial feature values ​​are assigned to each node and edge in the graph structure. The node features include local optical signal characteristics and environmental data, and the edge features include transmission loss and distance. Abnormality prediction module: The space-time matrix and graph structure are input into the pre-trained space-time graph convolutional network. The space-time graph convolutional network captures the dependencies between nodes and the dynamic changes in space and time in the graph structure, and predicts whether there are abnormalities in the optical fiber communication system. If there are abnormalities, the corresponding response measures are automatically executed and the alarm is triggered. The abnormal areas or nodes are marked in the graph structure and displayed to relevant personnel.

Images