A rail safety optical fiber intelligent early warning method and system

Abstract

The present invention discloses a rail safety optical fiber intelligent early warning method and system, comprising: collecting optical fiber vibration signals of tunnel optical fibers; performing vibration signal recognition on the optical fiber vibration signals to obtain vibration signal recognition results; and issuing early warnings based on the vibration signal recognition results. The present invention can perform real-time and uninterrupted monitoring of the target detection area to achieve intelligent monitoring and early warning.

Description

Technical Field

[0001] The present invention relates to the technical field of rail safety early warning, and in particular to a rail safety optical fiber intelligent early warning method and system. Background Art

[0002] The monitoring of the rail safety protection zone is of great significance to the protection of rail safety. In recent years, illegal and irregular construction in the rail safety protection zone has led to rail safety accidents. In order to solve this hidden danger, manual inspection methods have become the first choice. The characteristics of the manual inspection method are that it consumes a lot of manpower, material resources, and time to continuously inspect in the rail safety protection zone. Although this method has played a certain positive role in the monitoring of illegal and irregular construction, this method is laborious and ineffective and is not a long-term solution.

[0003] The drone inspection method is an upgrade of the manual inspection method. Compared with the traditional manual inspection, this method first liberates a large amount of labor and replaces human inspection with machine inspection. Then, this method combines artificial intelligence technology to intelligently identify whether there is construction behavior in the video taken by the drone, thereby improving the monitoring efficiency of the rail safety protection area.

[0004] Compared with manual inspection, drone inspection excels in technological innovation. However, both manual inspection and drone inspection have not jumped out of the logical framework of "inspection". Therefore, drone inspection still has the following shortcomings: 1) It is difficult to conduct 24-hour real-time monitoring of the rail safety protection zone; 2) It is difficult to conduct no-blind-angle monitoring of the rail safety protection zone; 3) It is difficult to conduct effective inspection of the rail safety protection zone at night; 4) It is greatly restricted by weather factors, and it is difficult to conduct effective monitoring of the rail safety protection zone in special weather such as fog, haze, wind, and rain. Summary of the invention

[0005] In order to solve the technical problems existing in the background technology, the present invention proposes a rail safety optical fiber intelligent early warning method and system.

[0006] The present invention proposes a rail safety optical fiber intelligent early warning method, comprising:

[0007] Collect optical fiber vibration signals of tunnel optical fibers;

[0008] Performing vibration signal recognition on the optical fiber vibration signal to obtain a vibration signal recognition result; wherein the vibration signal recognition result includes: the longitude of the vibration event, the latitude of the vibration event and the construction equipment;

[0009] Issue an early warning based on the vibration signal recognition results.

[0010] Preferably, collecting the optical fiber vibration signal of the tunnel optical fiber specifically includes:

[0011] Sending pulsed laser light into the tunnel fiber;

[0012] Capture the Rayleigh scattered light backscattered by the optical fiber;

[0013] Demodulate the captured Rayleigh scattered light;

[0014] The demodulated Rayleigh scattered light information is stored.

[0015] Preferably, before collecting the optical fiber vibration signal of the tunnel optical fiber, the method further comprises:

[0016] The correspondence between the length of the tunnel optical fiber, the tunnel position and the surface position is preset.

[0017] Preferably, the corresponding relationship between the length of the tunnel optical fiber, the tunnel position and the surface position is preset, specifically including:

[0018] Perform optical fiber tapping on the tunnel optical fiber in the tunnel to obtain a tapping signal response;

[0019] According to the knocking signal response, the length of the tunnel fiber is obtained;

[0020] Match different locations in the tunnel with the length of the tunnel fiber;

[0021] Obtain the design drawings of the tunnel and the sky map of the surface of the target monitoring area;

[0022] According to the design drawing of the tunnel and the sky map of the surface of the target monitoring area, different positions in the tunnel are matched with the surface positions to obtain the corresponding relationship between the length of the tunnel optical fiber, the tunnel position and the surface position.

[0023] Preferably, before collecting the optical fiber vibration signal of the tunnel optical fiber, the method further comprises:

[0024] Preset vibration events and corresponding optical fiber vibration signals.

[0025] The preset vibration events and corresponding optical fiber vibration signals specifically include:

[0026] Place a camera at the target location to record vibration events in real time;

[0027] When a vibration event occurs, vibration signals of different vibration events are extracted according to the corresponding relationship between the preset vibration event and the optical fiber vibration signal;

[0028] The preset correspondence between the vibration event and the optical fiber vibration signal is as follows: when the vibration event occurs, the corresponding optical fiber vibration signal appears; when the vibration event disappears, the corresponding optical fiber vibration signal disappears.

[0029] Preferably, performing vibration signal recognition on the optical fiber vibration signal to obtain a vibration signal recognition result specifically includes:

[0030] Based on geophysical methods, the vibration event signal characteristics are extracted from the time domain, frequency domain, space domain, FK domain and energy domain of the optical fiber vibration signal;

[0031] The pre-trained artificial intelligence recognition model is used to identify the characteristics of the vibration event signal to obtain the vibration signal recognition result.

[0032] Preferably, the artificial intelligence recognition model includes a first sub-model and a second sub-model, the first sub-model is based on a random forest algorithm and a convolutional neural network algorithm, and the second sub-model is based on a source scanning algorithm. The artificial intelligence recognition model is used to obtain the location of the vibration event signal and the type of construction equipment based on the first sub-model and the second sub-model.

[0033] Preferably, the vibration event signal features are identified using a pre-trained artificial intelligence recognition model to obtain a vibration signal recognition result, which specifically includes:

[0034] First, the pre-trained artificial intelligence recognition model is used to automatically identify the characteristics of the vibration event signal according to a preset period to obtain multiple recognition results; wherein the recognition results include the length of the optical fiber, the lateral distance of the vibration event signal from the tunnel, and the construction equipment;

[0035] When the optical fiber lengths corresponding to different recognition results differ within a preset distance, multiple recognition results are merged into one recognition result;

[0036] The optical fiber length of the fused recognition result is the average value of the optical fiber lengths in the multiple recognition results, and the lateral distance from the vibration event signal of the fused recognition result to the tunnel is the minimum value of the lateral distance from the vibration event signal in the multiple recognition results to the tunnel;

[0037] The optical fiber length corresponding to the fused recognition result and the lateral distance of the vibration event signal from the tunnel are converted into the longitude and latitude of the vibration event according to the coordinate system conversion method.

[0038] The present invention also proposes a rail safety optical fiber intelligent early warning system, the rail safety optical fiber intelligent early warning system includes: a data acquisition module, a signal intelligent recognition module and a visualization module, the data acquisition module is used to connect with the tunnel optical fiber, the data acquisition module, the signal intelligent recognition module and the visualization module are connected in sequence;

[0039] Wherein, the data acquisition module is used to collect optical fiber vibration signals of the tunnel optical fiber;

[0040] The signal intelligent recognition module is used to perform vibration signal recognition on the optical fiber vibration signal to obtain a vibration signal recognition result; wherein the vibration signal recognition result includes: the longitude of the vibration event, the latitude of the vibration event and the construction equipment;

[0041] The early warning module is used to issue an early warning based on the vibration signal recognition result.

[0042] Preferably, the data acquisition module comprises: a laser emission submodule, a weak scattered light capture submodule, a Rayleigh scattered light demodulation submodule and a data storage submodule;

[0043] The laser emission submodule is connected to the tunnel optical fiber, the laser emission submodule is connected to the weak scattered light capture submodule, the weak scattered light capture submodule is connected to the Rayleigh scattered light demodulation submodule, the Rayleigh scattered light demodulation module is connected to the data storage submodule, and the data storage submodule is connected to the signal intelligent recognition module;

[0044] The laser emission submodule is used for emitting pulsed laser into the tunnel optical fiber;

[0045] The weak scattered light capture submodule is used to capture the Rayleigh scattered light backscattered by the optical fiber;

[0046] The Rayleigh scattered light demodulation submodule is used to demodulate the captured Rayleigh scattered light;

[0047] The data storage submodule is used to store the demodulated Rayleigh scattered light information.

[0048] Preferably, the signal intelligent identification module includes a geophysical analysis submodule and an artificial intelligence analysis submodule;

[0049] The geophysical analysis submodule is used to extract the vibration event signal characteristics from the time domain, frequency domain, space domain, FK domain and energy domain of the optical fiber vibration signal based on geophysical methods to enhance the understanding of different types of signal characteristics;

[0050] The artificial intelligence analysis submodule is preset with a trained artificial intelligence recognition model. The artificial intelligence analysis submodule is used to use the artificial intelligence recognition model to identify the characteristics of the vibration event signal to obtain the vibration signal recognition result.

[0051] Preferably, the artificial intelligence recognition model includes a first sub-model and a second sub-model, the first sub-model is based on a random forest algorithm and a convolutional neural network (CNN) algorithm, and the second sub-model is based on a source scanning algorithm, and the artificial intelligence recognition model is used to obtain the location of the vibration event signal and the type of construction equipment based on the first sub-model and the second sub-model.

[0052] In the present invention, the proposed rail safety optical fiber intelligent early warning method and system first collects the optical fiber vibration signal of the existing tunnel optical fiber in the tunnel, and performs vibration signal recognition on the optical fiber vibration signal, and performs 7*24 hours real-time and uninterrupted monitoring of the target detection area according to the vibration signal recognition result to realize intelligent monitoring and early warning, without performing any processing on the existing optical fiber in the tunnel, and is immune to subway signal interference, and is not affected by special weather such as fog, haze, and strong wind, and has barrier-free monitoring at night, which effectively improves the monitoring efficiency of the rail safety protection zone, and reduces costs and increases efficiency for the maintenance and operation of the rail safety protection zone. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 The figure is a flow chart of a rail safety optical fiber intelligent early warning method in one embodiment of the present invention.

[0054] Figure 2 It is the optical fiber vibration signal corresponding to various vibration events in one embodiment of the present invention.

[0055] Figure 3 A characteristic diagram of a vibration event signal in an embodiment of the present invention.

[0056] Figure 4 This is a corresponding relationship diagram of the length of the tunnel optical fiber, the tunnel position and the surface position in one embodiment of the present invention. DETAILED DESCRIPTION

[0057] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0058] The present invention proposes a rail safety optical fiber intelligent early warning system, comprising: a data acquisition module, a signal intelligent recognition module and a visualization module, wherein the data acquisition module is used to connect with the tunnel optical fiber, and the data acquisition module, the signal intelligent recognition module and the visualization module are connected in sequence;

[0059] Wherein, the data acquisition module is used to collect optical fiber vibration signals of the tunnel optical fiber;

[0060] The signal intelligent recognition module is used to perform vibration signal recognition on the optical fiber vibration signal to obtain a vibration signal recognition result;

[0061] The early warning module is used to issue an early warning based on the vibration signal recognition result.

[0062] The present invention first collects the optical fiber vibration signal of the existing optical fiber in the tunnel (tunnel optical fiber), and performs vibration signal recognition on the optical fiber vibration signal, and realizes intelligent monitoring and early warning according to the vibration signal recognition result obtained by recognition, without any processing of the optical cable; it can perform 7 24-hour real-time uninterrupted monitoring, immune to subway signal interference, not affected by special weather conditions such as fog, haze, strong winds, and barrier-free monitoring at night, effectively improves the monitoring efficiency of the rail safety protection zone, and reduces costs and increases efficiency for the maintenance and operation of the rail safety protection zone.

[0063] The tunnel optical fiber in this embodiment is the optical fiber originally laid in the tunnel. There is no need to add or use new optical fibers, and there is zero impact on subway construction and operation.

[0064] In this embodiment, the data acquisition module includes: a laser emission submodule, a weak scattered light capture submodule, a Rayleigh scattered light demodulation submodule and a data storage submodule;

[0065] The laser emission submodule is connected to the tunnel optical fiber, the laser emission submodule is connected to the weak scattered light capture submodule, the weak scattered light capture submodule is connected to the Rayleigh scattered light demodulation submodule, the Rayleigh scattered light demodulation module is connected to the data storage submodule; the data storage submodule is connected to the signal intelligent recognition module;

[0066] The laser emission submodule is used for emitting pulsed laser into the tunnel optical fiber;

[0067] The weak scattered light capture submodule is used to capture the Rayleigh scattered light backscattered by the optical fiber;

[0068] The Rayleigh scattered light demodulation submodule is used to demodulate the captured Rayleigh scattered light;

[0069] The data storage submodule is used to store the demodulated Rayleigh scattered light information;

[0070] After the rail safety fiber optic intelligent warning system is assembled and debugged and can work normally, in order to obtain the length, tunnel position and surface position of the tunnel optical fiber that generates the fiber optic vibration signal, in this embodiment, the signal intelligent recognition module is preset with a correspondence between the optical fiber length, tunnel position and surface position.

[0071] First, the different locations in the tunnel are matched to the fiber lengths by tapping the fiber in the tunnel. Then, the design drawing of the tunnel is obtained and mapped to the sky map of the surface of the target detection area. In this way, different tunnel locations and surface locations are matched to each other. Therefore, the lengths of different tunnel fibers are matched to the tunnel locations and surface locations, such as Figure 4 As shown. Therefore, when a vibration event occurs at a certain position on the ground, the optical fiber length corresponding to the vibration event can be obtained, and then the corresponding optical fiber vibration signal can be obtained.

[0072] After the acquisition module is debugged and operates normally, in order to be able to distinguish the optical fiber vibration signals of various vibration events, in a further embodiment, the signal intelligent recognition module is preset with vibration events and corresponding optical fiber vibration signals.

[0073] Specifically, a camera is placed at a target location to record vibration events in real time, and the vibration signal of a vibration event is extracted based on the preset correspondence between vibration events and optical fiber vibration signals: "when a vibration event occurs, the corresponding optical fiber vibration signal appears; when the vibration event disappears, the corresponding optical fiber vibration signal disappears." Figure 2 shown.

[0074] Since there are many types of vibration signals, in order to intelligently identify the target construction signal from different event signals, in this embodiment, the signal intelligent identification module includes a geophysical analysis submodule and an artificial intelligence analysis submodule;

[0075] The geophysical analysis submodule is used to extract the vibration event signal characteristics from the time domain, frequency domain, space domain, FK domain and energy domain of the optical fiber vibration signal based on geophysical methods to enhance the understanding of different types of signal characteristics;

[0076] The artificial intelligence analysis submodule is preset with a trained artificial intelligence recognition model. The artificial intelligence analysis submodule is used to use the artificial intelligence recognition model to identify the characteristics of the vibration event signal to obtain the vibration signal recognition result.

[0077] Specifically, the artificial intelligence recognition model includes a first sub-model and a second sub-model. The first sub-model is based on a random forest algorithm and a convolutional neural network (CNN) algorithm, and the second sub-model is based on a source scanning algorithm.

[0078] The artificial intelligence recognition model is used to obtain the location and type of construction equipment of the vibration event signal based on the first sub-model and the second sub-model.

[0079] In one specific embodiment, the vibration event signal characteristics of roller construction, excavator construction, and steel sheet pile driver construction are as follows: Figure 3 shown.

[0080] Among them, the artificial intelligence analysis submodule uses historical optical fiber vibration signals and corresponding optical fiber vibration signal recognition results to train the artificial intelligence recognition model to improve the recognition accuracy of the artificial intelligence recognition model.

[0081] In this embodiment, the early warning module further includes a visualization submodule, and the visualization submodule is used to display the vibration signal recognition result and / or the text description of the early warning.

[0082] The visualization submodule is also used to display one or more of a map, a real-life image, and a construction site monitoring video.

[0083] Therefore, the early warning module can also be displayed in the visualization module through different early warning forms such as flashing red dots and shining five-pointed stars, and the warning duration and the early warning history information backtracking function can also be set.

[0084] Reference Figure 1 The present invention also proposes a rail safety optical fiber intelligent early warning method, comprising:

[0085] Collect optical fiber vibration signals of tunnel optical fibers;

[0086] Performing vibration signal recognition on the optical fiber vibration signal to obtain a vibration signal recognition result;

[0087] Issue an early warning based on the vibration signal recognition results.

[0088] The present invention collects the optical fiber vibration signal of the existing optical fiber in the tunnel (tunnel optical fiber), performs vibration signal recognition on the optical fiber vibration signal, and realizes intelligent monitoring and early warning according to the vibration signal recognition result obtained by recognition, without any processing of the optical cable; it can perform 7 24-hour real-time uninterrupted monitoring, immune to subway signal interference, not affected by special weather conditions such as fog, haze, strong winds, and barrier-free monitoring at night, effectively improves the monitoring efficiency of the rail safety protection zone, and reduces costs and increases efficiency for the maintenance and operation of the rail safety protection zone.

[0089] In order to accurately collect the optical fiber vibration signal, after the acquisition module is debugged and operates normally, in this embodiment, the optical fiber vibration signal of the tunnel optical fiber is collected, specifically including:

[0090] Sending pulsed laser light into the tunnel fiber;

[0091] Capture the Rayleigh scattered light backscattered by the optical fiber;

[0092] Demodulate the captured Rayleigh scattered light;

[0093] The demodulated Rayleigh scattered light information is stored.

[0094] After the acquisition module is debugged and operates normally, in order to obtain the length, tunnel position and surface position of the tunnel optical fiber that generates the optical fiber vibration signal, in this embodiment, before collecting the optical fiber vibration signal of the tunnel optical fiber, it also includes:

[0095] The correspondence between the length of the tunnel optical fiber, the tunnel position and the surface position is preset.

[0096] The correspondence between the preset tunnel fiber length, tunnel position and ground surface position specifically includes:

[0097] Perform optical fiber tapping on the tunnel optical fiber in the tunnel to obtain a tapping signal response;

[0098] According to the knocking signal response, the length of the tunnel fiber is obtained;

[0099] Match different locations in the tunnel with the length of the tunnel fiber;

[0100] Obtain the design drawings of the tunnel and the sky map of the surface of the target monitoring area;

[0101] According to the design drawing of the tunnel and the sky map of the surface of the target monitoring area, different positions in the tunnel are matched with the surface positions to obtain the corresponding relationship between the length of the tunnel optical fiber, the tunnel position and the surface position.

[0102] After the acquisition module is debugged and operates normally, in order to distinguish the optical fiber vibration signals of various vibration events, in this embodiment, before collecting the optical fiber vibration signal of the tunnel optical fiber, it also includes:

[0103] Preset vibration events and corresponding optical fiber vibration signals.

[0104] The preset vibration events and corresponding optical fiber vibration signals specifically include:

[0105] Place a camera at the target location to record vibration events in real time;

[0106] When a vibration event occurs, vibration signals of different vibration events are extracted according to the corresponding relationship between the preset vibration event and the optical fiber vibration signal;

[0107] The preset correspondence between the vibration event and the optical fiber vibration signal is as follows: when the vibration event occurs, the corresponding optical fiber vibration signal appears; when the vibration event disappears, the corresponding optical fiber vibration signal disappears.

[0108] In this embodiment, vibration signal recognition is performed on the optical fiber vibration signal to obtain a vibration signal recognition result, which specifically includes:

[0109] Based on geophysical methods, the vibration event signal characteristics are extracted from the time domain, frequency domain, space domain, FK domain and energy domain of the optical fiber vibration signal;

[0110] The artificial intelligence recognition model is used to identify the characteristics of vibration event signals and obtain vibration signal recognition results.

[0111] Among them, the artificial intelligence analysis submodule uses the length of the tunnel optical fiber, the correspondence between the tunnel position and the surface position, the vibration events and the corresponding optical fiber vibration signals, and the historical optical fiber vibration signals and the corresponding optical fiber vibration signal recognition results to train the artificial intelligence recognition model to improve the recognition accuracy of the artificial intelligence recognition model.

[0112] Among them, the vibration signal recognition results include: the longitude of the vibration event, the latitude of the vibration event and the type of construction equipment.

[0113] Among them, the artificial intelligence recognition model in this embodiment includes a first sub-model and a second sub-model. The first sub-model is based on a random forest algorithm and a convolutional neural network (CNN) algorithm, and the second sub-model is based on a source scanning algorithm. The artificial intelligence recognition model is used to obtain the location of the vibration event signal and the type of construction equipment based on the first sub-model and the second sub-model.

[0114] Therefore, in the recognition process, the pre-trained artificial intelligence recognition model is first used to automatically recognize the characteristics of the vibration event signal according to the preset period to obtain multiple recognition results; wherein the recognition results include the optical fiber length, the lateral distance of the vibration event signal from the tunnel, and the construction equipment;

[0115] When the optical fiber lengths corresponding to different recognition results differ within a preset distance, multiple recognition results are merged into one recognition result;

[0116] The optical fiber length of the fused recognition result is the average value of the optical fiber lengths in the multiple recognition results, and the lateral distance from the vibration event signal of the fused recognition result to the tunnel is the minimum value of the lateral distance from the vibration event signal in the multiple recognition results to the tunnel;

[0117] The optical fiber length corresponding to the fused recognition result and the lateral distance of the vibration event signal from the tunnel are converted into the longitude and latitude of the vibration event according to the coordinate system conversion method.

[0118] In one specific embodiment, the artificial intelligence recognition model including the first sub-model and the second sub-model is tested using a test set of optical fiber vibration signals of known construction machinery types. The test set of optical fiber vibration signals includes vibration signals and noise generated by excavators, crushers, road rollers, geological drills, sheet pile drivers, and subways. The test results are shown in the following table.

[0119] Missing reporting rate of construction monitoring and early warning: .

[0120] False alarm rate of construction monitoring and early warning: .

[0121] It can be seen that the present invention can greatly improve the accuracy of identification and reduce the missed alarm rate and false alarm rate of monitoring and early warning.

[0122] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A rail safety optical fiber intelligent early warning method, characterized in that: include: Collect optical fiber vibration signals of tunnel optical fibers; Performing vibration signal recognition on the optical fiber vibration signal to obtain a vibration signal recognition result; wherein the vibration signal recognition result includes: the longitude of the vibration event, the latitude of the vibration event and the construction equipment; Issue early warning based on vibration signal recognition results; Among them, the vibration signal recognition is performed on the optical fiber vibration signal to obtain the vibration signal recognition result, which specifically includes: Based on geophysical methods, the vibration event signal characteristics are extracted from the time domain, frequency domain, space domain, FK domain and energy domain of the optical fiber vibration signal; Use the pre-trained artificial intelligence recognition model to identify the characteristics of the vibration event signal and obtain the vibration signal recognition result; The artificial intelligence recognition model includes a first sub-model and a second sub-model, the first sub-model is based on a random forest algorithm and a convolutional neural network algorithm, the second sub-model is based on a source scanning algorithm, and the artificial intelligence recognition model is used to obtain the location of the vibration event signal and the type of construction equipment based on the first sub-model and the second sub-model; Among them, the vibration event signal characteristics are identified using a pre-trained artificial intelligence recognition model to obtain the vibration signal recognition results, which specifically include: First, the pre-trained artificial intelligence recognition model is used to automatically identify the characteristics of the vibration event signal according to a preset period to obtain multiple recognition results; wherein the recognition results include the length of the optical fiber, the lateral distance of the vibration event signal from the tunnel, and the construction equipment; When the optical fiber lengths corresponding to different recognition results differ within a preset distance, multiple recognition results are merged to form one recognition result; wherein the optical fiber length of the merged recognition result is the average value of the optical fiber lengths in the multiple recognition results, and the lateral distance of the vibration event signal from the tunnel in the merged recognition result is the minimum value of the lateral distance of the vibration event signal from the tunnel in the multiple recognition results; The optical fiber length corresponding to the fused recognition result and the lateral distance of the vibration event signal from the tunnel are converted into the longitude and latitude of the vibration event according to the coordinate system conversion method.

2. The rail safety optical fiber intelligent early warning method according to claim 1 is characterized in that: Collect the optical fiber vibration signal of the tunnel optical fiber, including: Sending pulsed laser light into the tunnel fiber; Capture the Rayleigh scattered light backscattered by the optical fiber; Demodulate the captured Rayleigh scattered light; The demodulated Rayleigh scattered light information is stored.

3. The rail safety optical fiber intelligent early warning method according to claim 1 is characterized in that: Before collecting the optical fiber vibration signal of the tunnel optical fiber, it also includes: Preset the correspondence between the length of tunnel optical fiber, tunnel position and surface position; The correspondence between the preset tunnel fiber length, tunnel position and ground surface position specifically includes: Perform optical fiber tapping on the tunnel optical fiber in the tunnel to obtain a tapping signal response; According to the knocking signal response, the length of the tunnel fiber is obtained; Match different locations in the tunnel with the length of the tunnel fiber; Obtain the design drawings of the tunnel and the sky map of the surface of the target monitoring area; According to the design drawing of the tunnel and the sky map of the surface of the target monitoring area, different positions in the tunnel are matched with the surface positions to obtain the corresponding relationship between the length of the tunnel optical fiber, the tunnel position and the surface position.

4. The rail safety optical fiber intelligent early warning method according to claim 3 is characterized in that: Before collecting the optical fiber vibration signal of the tunnel optical fiber, it also includes: Preset vibration events and corresponding optical fiber vibration signals; The preset vibration events and corresponding optical fiber vibration signals specifically include: Place a camera at the target location to record vibration events in real time; When a vibration event occurs, vibration signals of different vibration events are extracted according to the corresponding relationship between the preset vibration event and the optical fiber vibration signal; The preset correspondence between the vibration event and the optical fiber vibration signal is as follows: when the vibration event occurs, the corresponding optical fiber vibration signal appears; when the vibration event disappears, the corresponding optical fiber vibration signal disappears.

5. A rail safety optical fiber intelligent early warning system, characterized in that: The rail safety optical fiber intelligent early warning system comprises: a data acquisition module, a signal intelligent recognition module and a visualization module, wherein the data acquisition module is used to connect with the tunnel optical fiber, and the data acquisition module, the signal intelligent recognition module and the visualization module are connected in sequence; Wherein, the data acquisition module is used to collect optical fiber vibration signals of the tunnel optical fiber; The signal intelligent recognition module is used to perform vibration signal recognition on the optical fiber vibration signal to obtain a vibration signal recognition result; wherein the vibration signal recognition result includes: the longitude of the vibration event, the latitude of the vibration event and the construction equipment; An early warning module is used to issue an early warning based on the vibration signal recognition result; Among them, the signal intelligent identification module includes a geophysical analysis submodule and an artificial intelligence analysis submodule; The geophysical analysis submodule is used to extract the vibration event signal characteristics from the time domain, frequency domain, space domain, FK domain and energy domain of the optical fiber vibration signal based on geophysical methods; The artificial intelligence analysis submodule is preset with a trained artificial intelligence recognition model, and the artificial intelligence analysis submodule is used to use the artificial intelligence recognition model to identify the characteristics of the vibration event signal to obtain the vibration signal recognition result; wherein the artificial intelligence recognition model includes a first submodel and a second submodel, the first submodel is based on a random forest algorithm and a convolutional neural network algorithm, and the second submodel is based on a source scanning algorithm, and the artificial intelligence recognition model is used to obtain the location of the vibration event signal and the type of construction equipment based on the first submodel and the second submodel; During the recognition process, the pre-trained artificial intelligence recognition model is first used to automatically recognize the characteristics of the vibration event signal according to a preset period to obtain multiple recognition results; the recognition results include the length of the optical fiber, the lateral distance of the vibration event signal from the tunnel, and the construction equipment; When the optical fiber lengths corresponding to different recognition results differ within a preset distance, multiple recognition results are merged into one recognition result; The optical fiber length of the fused recognition result is the average value of the optical fiber lengths in the multiple recognition results, and the lateral distance from the vibration event signal of the fused recognition result to the tunnel is the minimum value of the lateral distance from the vibration event signal in the multiple recognition results to the tunnel; The optical fiber length corresponding to the fused recognition result and the lateral distance of the vibration event signal from the tunnel are converted into the longitude and latitude of the vibration event according to the coordinate system conversion method.

6. The rail safety optical fiber intelligent early warning system according to claim 5 is characterized in that: The data acquisition module includes: a laser emission submodule, a weak scattered light capture submodule, a Rayleigh scattered light demodulation submodule and a data storage submodule; The laser emission submodule is connected to the tunnel optical fiber, the laser emission submodule is connected to the weak scattered light capture submodule, the weak scattered light capture submodule is connected to the Rayleigh scattered light demodulation submodule, the Rayleigh scattered light demodulation module is connected to the data storage submodule, and the data storage submodule is connected to the signal intelligent recognition module; The laser emission submodule is used to emit pulsed laser into the tunnel optical fiber; The weak scattered light capture submodule is used to capture the Rayleigh scattered light backscattered by the optical fiber; The Rayleigh scattered light demodulation submodule is used to demodulate the captured Rayleigh scattered light; The data storage submodule is used to store the demodulated Rayleigh scattered light information.

Images