Low-altitude dynamic fence monitoring method and system based on train positioning protection level
By using a dynamic fence monitoring method based on train positioning protection level, the parameters of the low-altitude monitoring terminal are adjusted in real time, which solves the problems of resource waste and positioning error in the existing technology and realizes accurate monitoring and efficient resource utilization for railway low-altitude safety.
Patent Information
- Application Number
- CN202610214119.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-02-13
- Publication Date
- 2026-03-20
AI Technical Summary
The existing railway low-altitude safety monitoring system lacks spatiotemporal dynamics, resulting in wasted resources and invalid alarms. Furthermore, it suffers from large positioning errors in complex electromagnetic environments, making it impossible to achieve precise vehicle-ground-air coordinated defense.
By introducing a train positioning protection level, the geometric dimensions of the dynamic electronic fence are calculated in real time. Combined with the train's motion status and satellite signal quality, the detection parameters of the low-altitude monitoring terminal are dynamically adjusted to construct a follow-up three-dimensional dynamic electronic fence, thereby achieving precise monitoring of low-altitude targets.
It effectively eliminates safety blind spots caused by positioning errors, reduces false alarm rate and resource consumption, and improves the detection efficiency of low-altitude targets in complex electromagnetic environments.
Smart Images

Figure CN121697705A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the technical field of intelligent transportation systems (ITS) and railway operation safety monitoring, and in particular relates to a low-altitude dynamic fence monitoring method and system based on train positioning protection level. Background Technology
[0002] With the continuous increase in the operating mileage of high-speed railways (HSR) and the explosive development of the "low-altitude economy," low-altitude safety along railway lines faces severe challenges. Incidents of consumer drones, agricultural and forestry drones, as well as kites, balloons, and other low-altitude floating objects intruding into railway clearances occur frequently, seriously threatening the safety of high-speed train operations.
[0003] Currently, surveillance technologies targeting railway perimeter intrusion have made some progress. The mainstream approach typically employs a "multi-sensor fusion" architecture, deploying millimeter-wave radar for wide-area detection in key sections along the railway line (such as stations, bridges, and tunnel entrances), and linking it with photoelectric cameras for target verification and tracking. To delineate alarm zones, existing technologies generally use "static geofencing," which defines a rectangular virtual no-fly zone extending a fixed distance (e.g., 500 meters to the left and right, and 200 meters high) from the railway track centerline. Once sensors detect a target entering this fixed geometric area, an alarm or countermeasures are triggered.
[0004] However, the aforementioned existing technologies have significant technical defects and limitations in practical applications, lacking "spatiotemporal dynamism," leading to resource waste and invalid alarms. Existing static electronic fences are effective all-time and all-airspace. Regardless of whether a train is passing through the current section, the monitoring system operates under high load. This not only results in huge energy consumption for radar and computing units but also generates a large number of invalid interference alarms. For example, when no train is running on the track, a bird or a drone operating at a safe distance does not pose a substantial threat, but the static fence will still trigger an alarm, causing train dispatchers or security personnel to face a serious "crying wolf" effect, reducing the actual efficiency of system response, ignoring the "integrity risk" of satellite positioning, and posing a risk of missed alarms. In collaborative monitoring scenarios involving train positions, existing attempts often directly use the positioning coordinates (x, y, z) output by the train's GNSS receiver to determine the relative position of the train and the drone. However, the environment along high-speed railways is extremely complex, with numerous mountain obstructions, tunnel entrances and exits, and electromagnetic environments from high-voltage contact networks. In these areas, GNSS signals are highly susceptible to interference from multipath effects and non-line-of-sight (NLOS) propagation, leading to positioning errors of tens or even hundreds of meters. Existing systems assume absolute accuracy in positioning coordinates and do not consider the statistical boundaries of positioning errors. If the actual position of the train exceeds the system's preset tolerance due to signal drift, and the monitoring system still generates a protection zone based on incorrect coordinates, the actual danger zone (the train's actual location) will lose monitoring coverage, creating a fatal safety blind spot.
[0005] Sensor resource scheduling lacks the ability to "perceive uncertainty." Current ground surveillance radars and cameras typically employ fixed scanning patterns or simple distance-based zoom logic. The system cannot perceive the "credibility" of the current train's location. In high-risk moments when train positioning errors increase (e.g., just exiting a tunnel), existing systems cannot automatically expand the search range (i.e., "expand" the surveillance area) to compensate for positioning errors; conversely, in moments of extremely precise positioning, the system cannot automatically narrow the surveillance range to focus on the core threat. This surveillance strategy, lacking adaptive flexibility, struggles to achieve precise "vehicle-ground-air" coordinated defense in complex electromagnetic environments.
[0006] In summary, there is an urgent need for a dynamic fence generation and monitoring method that can combine the real-time movement status of trains and fully consider the protection level of satellite positioning integrity, in order to solve the shortcomings of existing static fence technology in terms of responsiveness, anti-error interference capability, and resource scheduling efficiency. Summary of the Invention
[0007] In view of this, the present invention aims to overcome the shortcomings of the above-mentioned problems in the prior art and proposes a low-altitude dynamic fence monitoring method and system based on train positioning protection level. By introducing train positioning integrity parameters, a dynamic defense space that changes in real time with train position and signal quality is constructed, which solves the problems of accuracy and timeliness of low-altitude intrusion monitoring along railway lines in complex electromagnetic environments.
[0008] To achieve the above objectives, the technical solution of the present invention is implemented as follows:
[0009] In a first aspect, the present invention provides a low-altitude dynamic fence monitoring method based on train positioning protection level, comprising the following steps:
[0010] Step 1: By deploying a vehicle-mounted high-completeness positioning unit on the train, the current motion status data and positioning completeness parameters of the train are obtained in real time. The positioning completeness parameters include the protection level.
[0011] Step 2: Based on the protection level and the current motion state data, calculate the geometric dimension parameters of the dynamic electronic fence in real time, and construct a follow-up three-dimensional dynamic electronic fence with the current position of the train as the reference center;
[0012] Step 3: The spatial range data of the three-dimensional dynamic electronic fence is synchronously transmitted to the ground monitoring subsystem along the railway line via a wireless communication network;
[0013] Step 4: The ground monitoring subsystem dynamically adjusts the detection parameters of the connected low-altitude monitoring terminal based on the received spatial range data of the three-dimensional dynamic electronic fence, and generates monitoring alarm results for targets falling into the three-dimensional dynamic electronic fence.
[0014] Furthermore, in step 1, the protection level is the statistical upper bound of the position error calculated based on the receiver autonomous completeness monitoring algorithm or weighted least squares residual analysis, including the horizontal protection level and the vertical protection level.
[0015] Furthermore, the method also includes:
[0016] The system monitors the number of visible satellites and the geometric accuracy factor of the train's environment in real time. When the number of visible satellites decreases or the geometric accuracy factor deteriorates, causing the statistical upper bound of the positioning error to increase, the value of the protection level is automatically increased.
[0017] Furthermore, in step 2, constructing a follow-up three-dimensional dynamic electronic fence with the train's current position as the reference center includes:
[0018] Construct an ellipsoidal or polyhedral model centered at the train's positioning coordinates; the lateral radius of the electronic fence. The longitudinal radius of the electronic fence is positively correlated with the horizontal protection level. Based on the horizontal protection level and the train's current speed vector The calculation is performed using linear combinations; the specific calculation logic satisfies the following relationship:
[0019]
[0020]
[0021] Where k and α are weighting coefficients, and v is the current velocity along the track tangent. For vehicle-to-ground communication delay time, The safety redundancy constant preset by C is for system response time.
[0022] Furthermore, the low-altitude surveillance terminal includes a radar detection device, and the detection parameters of the connected low-altitude surveillance terminal are dynamically adjusted as follows:
[0023] The spatial range of the three-dimensional dynamic electronic fence is mapped to a spatiotemporal mask of radar;
[0024] For detected low-altitude echo targets, determine whether their coordinates are located inside the spatiotemporal mask;
[0025] If the target is located inside the spatiotemporal mask, the constant false alarm rate detection threshold or the velocity threshold of the moving target indication filter in that area is automatically reduced to improve the detection rate of targets with small radar cross-sections.
[0026] Furthermore, the low-altitude monitoring terminal includes a camera device with optical zoom function, and the detection parameters of the connected low-altitude monitoring terminal are dynamically adjusted, including:
[0027] Establish camera focal length With the horizontal radius of the electronic fence The inverse mapping relationship between them;
[0028] When the protection level value increases, it causes the horizontal radius of the electronic fence to increase. During expansion, the camera is controlled to perform a wide-angle zoom operation to expand the field of view;
[0029] When the protection level value decreases, it causes the horizontal radius of the electronic fence to decrease. When retracting, control the camera to perform a telephoto zoom operation to provide close-up monitoring of the fenced area.
[0030] Furthermore, the method also includes a processing step for satellite navigation signal rejection:
[0031] When the onboard high-completeness positioning unit detects that the satellite signal loss time exceeds a preset threshold, it automatically switches to the inertial navigation calculation mode. In the inertial navigation calculation mode, the value of the protection level increases quadratically or exponentially with time. The spatial volume of the three-dimensional dynamic electronic fence expands rapidly until it covers the area of greatest uncertainty that the train may have.
[0032] Furthermore, the method also includes a hierarchical monitoring response step:
[0033] For low-altitude targets entering the three-dimensional dynamic electronic fence, the system determines their attributes: if it is a registered cooperative target, the boundary data of the three-dimensional dynamic electronic fence is pushed to the target's flight control system in real time to trigger its automatic avoidance logic; if it is a non-cooperative target, the collision time is calculated based on the relative distance between the target and the electronic fence boundary, and an audible and visual alarm or electromagnetic interference countermeasures are triggered.
[0034] Secondly, the present invention provides a low-altitude dynamic fence monitoring system based on train positioning protection level, comprising:
[0035] The vehicle-mounted subsystem includes a high-completeness positioning unit and a vehicle-mounted communication module; the high-completeness positioning unit is configured to perform a step of acquiring a protection level and to send data through the vehicle-mounted communication module;
[0036] The ground surveillance subsystem includes a data processing center and multiple low-altitude surveillance terminals deployed along the railway line; the data processing center is configured to receive data sent by the vehicle-mounted subsystem and perform the step of constructing a three-dimensional dynamic electronic fence; the low-altitude surveillance terminals are configured to receive the spatial range data of the three-dimensional dynamic electronic fence and perform the step of dynamically adjusting the detection parameters.
[0037] Compared with existing technologies, the low-altitude dynamic fence monitoring method and system based on train positioning protection level described in this invention has the following advantages:
[0038] This invention can eliminate safety blind spots caused by positioning errors. When the positioning error increases due to poor satellite signal (such as in mountainous areas or tunnel entrances), the system automatically "expands" the fence radius through a formula, using space to exchange for safety, and completely solves the problem of missed reports caused by positioning drift.
[0039] This invention significantly reduces false alarm rates and resource consumption. Compared to static fences that are always open, the dynamic fence of this invention moves with the vehicle. This greatly reduces invalid alarms triggered by irrelevant background objects and minimizes interference with security personnel.
[0040] This invention enables intelligent scheduling of sensor resources and establishes a mathematical mapping relationship between radar / photoelectric parameters and train protection levels. When uncertainty is high, it automatically widens the angle and lowers the threshold for a "safety net" search; when certainty is high, it focuses on close-ups, significantly improving the detection efficiency of low-altitude small targets in complex electromagnetic environments. Attached Figure Description
[0041] The accompanying drawings, which form part of this invention, 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 undue limitation of the invention. In the drawings:
[0042] Figure 1 This is a schematic diagram of a low-altitude dynamic fence monitoring method based on train positioning protection level according to the present invention.
[0043] Figure 2 This is a schematic diagram of a low-altitude dynamic fence monitoring system based on train positioning protection level according to the present invention. Detailed Implementation
[0044] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.
[0045] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.
[0046] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.
[0047] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.
[0048] Example 1
[0049] like Figure 1 As shown, this invention provides a low-altitude dynamic fence monitoring method based on train positioning protection level, comprising the following steps:
[0050] Step 1: Obtain the train's current motion status data and positioning integrity parameters in real time through the on-board high integrity positioning unit, where the positioning integrity parameters include at least the Protection Level (PL).
[0051] Step 2: Based on the protection level (PL) and the current motion state data of the train, calculate the spatial geometric parameters of the dynamic electronic fence in real time, and construct a follow-up three-dimensional dynamic electronic fence with the current position of the train as the reference center.
[0052] Step 3: Synchronously transmit the spatial coordinate data of the 3D dynamic electronic fence to the low-altitude monitoring terminal along the railway line;
[0053] Step 4: The low-altitude monitoring terminal dynamically adjusts the detection strategy or parameters for low-altitude targets based on the received three-dimensional dynamic electronic fence range, and generates monitoring results for targets falling within the fence.
[0054] As a further step, in step 1, the train's onboard high-completeness positioning unit (such as a high-precision GNSS receiver) acquires the train's motion status data and positioning completeness parameters (Protection Level, PL) in real time. PL includes the Horizontal Protection Level (HPL) and the Vertical Protection Level (VPL). These parameters are calculated using the Receiver Autonomous Completeness Monitoring (RAIM) algorithm or Weighted Least Squares (WLS) residual analysis, representing the statistical upper bound of the positioning error.
[0055] The RAIM algorithm calculates PL based on the residuals between the received observations and the theoretical values using the following formula:
[0056]
[0057] in, Let σ be the residual of the i-th observation, σ be the standard deviation of the corresponding residual, n be the number of observation epochs, and Threshold be the preset error threshold. This formula is used to estimate the upper limit of the positioning error and dynamically adjust the PL value according to changes in the environment.
[0058] As a further solution, in step 2, the specific method for constructing a responsive 3D dynamic electronic fence is as follows: construct an ellipsoid or polyhedron model centered on the train's positioning coordinates; the horizontal radius of the electronic fence... It is positively correlated with the horizontal protection level (HPL); the longitudinal radius of the electronic fence Based on the horizontal protection level (HPL) and the train's current speed vector The calculation is performed using linear combinations; the specific calculation logic satisfies the following relationship:
[0059]
[0060]
[0061] Where k and α are weighting coefficients, and v is the current velocity along the track tangent. For vehicle-to-ground communication delay time, The safety redundancy constant preset by C is for system response time.
[0062] At this point, the system constructs a three-dimensional dynamic electronic fence centered on the train's current position. Based on the train's speed and PL (Power Line Shift), the system dynamically adjusts the fence's radius to accommodate different motion states and satellite signal quality. The volume of the three-dimensional dynamic electronic fence is expressed by the following formula:
[0063]
[0064] in, The volume and lateral radius of the three-dimensional dynamic electronic fence. Longitudinal radius .
[0065] As a further step, in step 4, the low-altitude monitoring terminal includes millimeter-wave radar equipment. When dynamically adjusting the detection strategy or parameters for low-altitude targets, the onboard system transmits the train's current position and geofence data to the ground monitoring subsystem in real time via a wireless communication network. This data includes the coordinate data of the 3D geofence, the current train speed, acceleration, and positioning completeness parameters. After receiving the geofence data, the ground monitoring terminal dynamically adjusts its detection parameters based on the train's current motion state and PL value. Specifically, the radar system adjusts its spatiotemporal mask to optimize detection efficiency. For example:
[0066]
[0067] in, and To adjust the coefficient, and To adjust the constant false alarm rate before and after, the system can flexibly respond to various targets in different security environments by dynamically adjusting the CFAR.
[0068] As a further solution, the low-altitude surveillance terminal includes a camera device with optical zoom capability; when dynamically adjusting the detection strategy or parameters for low-altitude targets, the relationship between the camera focal length f and the horizontal radius of the electronic fence is established. The inverse mapping relationship between them; when the protection level (PL) value increases, the horizontal radius of the electronic fence increases. During expansion, the camera performs a wide-angle zoom operation to increase the field of view (FOV); when the protection level (PL) value decreases, the horizontal radius of the electronic fence decreases. When retracting, control the camera to perform a telephoto zoom operation to provide close-up monitoring of the fenced area.
[0069] As a further solution, it also includes processing steps for satellite navigation signal rejection: when the onboard high-completeness positioning unit detects that the satellite signal loss duration exceeds a preset threshold, it automatically switches to inertial navigation estimation mode; in inertial navigation estimation mode, the protection level (PL) value increases quadratically or exponentially with time, specifically as follows:
[0070]
[0071] Where PL(0) is the initial protection level, ε is the growth coefficient, and t is time. In this mode, the system expands the radius of the electronic fence to cope with the area of maximum possible positioning uncertainty.
[0072] The spatial volume of the three-dimensional dynamic electronic fence expands rapidly until it covers the area of greatest uncertainty that the train may present.
[0073] As a further solution, the method also includes a tiered monitoring and response step: For low-altitude targets entering the 3D dynamic electronic fence, the system determines their attributes: if it is a registered cooperative target, the boundary data of the 3D dynamic electronic fence is pushed to the target's flight control system in real time, triggering its automatic avoidance logic; if it is a non-cooperative target, the collision time (TTC) is calculated based on the relative distance between the target and the electronic fence boundary, specifically:
[0074]
[0075] in, It is the distance between the target and the fence boundary. It is the speed of the target.
[0076] If the TTC value falls below a certain set threshold, the system will trigger an alarm to alert security personnel or dispatchers to take appropriate measures. For high-risk targets, the system can also activate electromagnetic interference countermeasures.
[0077] Example 2
[0078] like Figure 2As shown, a low-altitude dynamic fence monitoring system based on train positioning protection includes:
[0079] The vehicle-mounted subsystem includes a high-integrity positioning unit and a vehicle-mounted communication module; the high-integrity positioning unit is configured to perform protection-level steps and transmit data via the vehicle-mounted communication module;
[0080] The ground surveillance subsystem includes a data processing center and multiple low-altitude surveillance terminals deployed along the railway line; the data processing center is configured to receive data sent by the vehicle-mounted subsystem and perform the step of constructing a three-dimensional dynamic electronic fence; the low-altitude surveillance terminals are configured to receive the spatial range data of the three-dimensional dynamic electronic fence and perform the step of dynamically adjusting the detection parameters.
[0081] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A low-altitude dynamic fence monitoring method based on train positioning protection level, characterized in that: Includes the following steps: Step 1: By deploying a vehicle-mounted high-completeness positioning unit on the train, the current motion status data and positioning completeness parameters of the train are obtained in real time. The positioning completeness parameters include the protection level. Step 2: Based on the protection level and the current motion state data, calculate the geometric dimension parameters of the dynamic electronic fence in real time, and construct a follow-up three-dimensional dynamic electronic fence with the current position of the train as the reference center; Step 3: The spatial range data of the three-dimensional dynamic electronic fence is synchronously transmitted to the ground monitoring subsystem along the railway line via a wireless communication network; Step 4: The ground monitoring subsystem dynamically adjusts the detection parameters of the connected low-altitude monitoring terminal based on the received spatial range data of the three-dimensional dynamic electronic fence, and generates monitoring alarm results for targets falling into the three-dimensional dynamic electronic fence.
2. The low-altitude dynamic fence monitoring method based on train positioning protection level according to claim 1, characterized in that: In step 1, the protection level is the upper bound of the position error statistical calculation based on the receiver autonomous completeness monitoring algorithm or weighted least squares residual analysis, which includes the horizontal protection level and the vertical protection level.
3. The low-altitude dynamic fence monitoring method based on train positioning protection level according to claim 1, characterized in that: The method further includes: The system monitors the number of visible satellites and the geometric accuracy factor of the train's environment in real time. When the number of visible satellites decreases or the geometric accuracy factor deteriorates, causing the statistical upper bound of the positioning error to increase, the value of the protection level is automatically increased.
4. The low-altitude dynamic fence monitoring method based on train positioning protection level according to claim 1, characterized in that: In step 2, constructing a homing three-dimensional dynamic electronic fence with the train's current position as the reference center includes: Construct an ellipsoidal or polyhedral model centered at the train's positioning coordinates; the lateral radius of the electronic fence. The longitudinal radius of the electronic fence is positively correlated with the horizontal protection level. Based on the horizontal protection level and the train's current speed vector The calculation is performed using linear combinations; the specific calculation logic satisfies the following relationship: ; ; Where k and α are weighting coefficients, and v is the current velocity along the track tangent. For vehicle-to-ground communication delay time, The safety redundancy constant preset by C is for system response time.
5. A low-altitude dynamic fence monitoring method based on train positioning protection level according to claim 1, characterized in that: The low-altitude surveillance terminal includes a radar detection device, and the detection parameters of the connected low-altitude surveillance terminal are dynamically adjusted as follows: The spatial range of the three-dimensional dynamic electronic fence is mapped to a spatiotemporal mask of radar; For detected low-altitude echo targets, determine whether their coordinates are located inside the spatiotemporal mask; If the target is located inside the spatiotemporal mask, the constant false alarm rate detection threshold or the velocity threshold of the moving target indication filter in that area is automatically reduced to improve the detection rate of targets with small radar cross-sections.
6. A low-altitude dynamic fence monitoring method based on train positioning protection level according to claim 1, characterized in that: The low-altitude monitoring terminal includes a camera device with optical zoom function, and the detection parameters of the connected low-altitude monitoring terminal are dynamically adjusted as follows: Establish the camera focal length f and the horizontal radius of the electronic fence. The inverse mapping relationship between them; When the protection level value increases, it causes the horizontal radius of the electronic fence to increase. During expansion, the camera is controlled to perform a wide-angle zoom operation to expand the field of view; When the protection level value decreases, it causes the horizontal radius of the electronic fence to decrease. When retracting, control the camera to perform a telephoto zoom operation to provide close-up monitoring of the fenced area.
7. A low-altitude dynamic fence monitoring method based on train positioning protection level according to claim 1, characterized in that: The method also includes processing steps for satellite navigation signal rejection scenarios: When the onboard high-completeness positioning unit detects that the satellite signal loss time exceeds a preset threshold, it automatically switches to the inertial navigation calculation mode. In the inertial navigation calculation mode, the value of the protection level increases quadratically or exponentially with time. The spatial volume of the three-dimensional dynamic electronic fence expands rapidly until it covers the area of greatest uncertainty that the train may have.
8. A low-altitude dynamic fence monitoring method based on train positioning protection level according to claim 1, characterized in that: The method also includes a hierarchical monitoring response step: For low-altitude targets entering the three-dimensional dynamic electronic fence, the system determines their attributes: if it is a registered cooperative target, the boundary data of the three-dimensional dynamic electronic fence is pushed to the target's flight control system in real time to trigger its automatic avoidance logic. If the target is not cooperative, the collision time is calculated based on the relative distance between the target and the electronic fence boundary, and an audible and visual alarm or electromagnetic interference countermeasures are triggered.
9. A low-altitude dynamic fence monitoring system based on train positioning protection level, characterized in that: It includes an onboard subsystem, a high-integrity positioning unit, and an onboard communication module; the high-integrity positioning unit is configured to perform a step of acquiring a protection level and to send data through the onboard communication module; The ground surveillance subsystem includes a data processing center and multiple low-altitude surveillance terminals deployed along the railway line; the data processing center is configured to receive data sent by the vehicle-mounted subsystem and perform the step of constructing a three-dimensional dynamic electronic fence. The low-altitude monitoring terminal is configured to receive the spatial range data of the three-dimensional dynamic electronic fence and perform the step of dynamically adjusting the detection parameters.