Perimeter intrusion monitoring method based on optical fiber sensing technology

Abstract

The invention relates to the technical field of perimeter security monitoring, and discloses a perimeter intrusion monitoring method based on an optical fiber sensing technology, and the method comprises the steps: deploying a main sensing optical cable according to a perimeter contour to form a first monitoring surface and a second monitoring surface which is not coplanar with the first monitoring surface, and enabling a vibration optical fiber host to emit a laser signal along the sensing optical cable, the sensing optical cable converts a vibration signal into scattered return light and transmits the scattered return light back to the vibration optical fiber host, the vibration optical fiber host sequentially performs photoelectric conversion processing and acquisition analysis on the returned scattered return light to finally obtain a suspected intrusion signal, and a suspected intrusion grading signal set in a specified time period is obtained in an equal period. Based on core data of a suspected intrusion grading signal set, an AI algorithm can be used for generating an intrusion fitting track after suspected intrusion point location grouping, the intrusion fitting track is judged, effective intrusion behaviors are judged and positioned, and an intrusion grading alarm result is obtained.

Description

Technical Field

[0001] This application relates to the technical field of perimeter security monitoring, and in particular to a perimeter intrusion monitoring method based on fiber optic sensing technology. Background Technology

[0002] In the field of perimeter security monitoring, traditional intrusion detection technologies often rely on a single sensing method or a single monitoring surface, which has many limitations. Single-sensor technologies have weak anti-interference capabilities and are easily affected by electromagnetic interference and environmental noise, resulting in a high false alarm rate and difficulty in accurately distinguishing between real intrusions and environmental interference.

[0003] Traditional monitoring methods often employ a single-plane monitoring layout, which cannot comprehensively capture the three-dimensional movement trajectory of intrusion targets. For intrusion behaviors that cross different monitoring areas and exhibit continuous movement characteristics, it is difficult to achieve complete trajectory tracking and correlation analysis, often resulting in monitoring gaps or missed reports. Furthermore, existing technologies lack a scientific hierarchical early warning mechanism, failing to provide differentiated warnings based on the degree of danger of intrusion behavior. This makes it difficult for security personnel to quickly focus on high-risk events, leading to low emergency response efficiency.

[0004] Therefore, there is an urgent need for a multi-dimensional, high-precision, and interference-resistant perimeter intrusion detection method to meet the diverse security needs of different scenarios. Summary of the Invention

[0005] The purpose of this invention is to provide a perimeter intrusion detection method based on fiber optic sensing technology, thereby solving at least one of the above-mentioned problems.

[0006] This invention provides a perimeter intrusion detection method based on fiber optic sensing technology, the method comprising: Step S1: Deploy the main sensing optical cable according to the perimeter contour to form a first monitoring surface, and draw out the secondary sensing optical cable from the main sensing optical cable to form a second monitoring surface that is not coplanar with the first monitoring surface; Step S2: Process the vibration signal of the sensing optical cable and obtain a set of suspected intrusion classification signals within a specified time period at equal intervals. Step S3: Based on the core data of the suspected intrusion classification signal set, suspected intrusion points are grouped and intrusion fitting trajectories are generated; Step S4: Determine the intrusion fitting trajectory, identify and locate valid intrusion behaviors, and obtain intrusion classification alarm results.

[0007] As a further technical solution, the method for processing the vibration signal of the sensing optical cable includes: The vibration fiber optic host emits a laser signal along the sensing optical cable. When there is external vibration, the sensing optical cable converts the vibration signal into scattered light and transmits it back to the vibration fiber optic host. The vibration fiber optic host sequentially performs photoelectric conversion processing on the returned scattered light and collects and analyzes the vibration signal to finally obtain the suspected intrusion signal.

[0008] As a further technical solution, the method for obtaining the suspected intrusion classification signal set includes: The second monitoring surface is divided into multiple sector-shaped monitoring areas, and key parameters of the sector-shaped monitoring areas are configured, including the center position, sector angle, and interval distance. The level m of the corresponding division region is determined based on the number of overlaps of each division region in the sector-shaped monitoring area; For each sector-shaped monitoring area, the set of suspected intrusion signals obtained by the vibration fiber optic host is taken as the m-level suspected intrusion point set; the set of suspected intrusion classification signals is the set of the m-level suspected intrusion point set.

[0009] As a further technical solution, the method for grouping suspected intrusion points includes: For each set of m-level suspected intrusion points, the corresponding m-level suspected intrusion point metadata is confirmed; the suspected intrusion point metadata includes the suspected intrusion start coordinates and timestamp; Establish the i-th data group containing the metadata of the m-level suspected intrusion points; Based on specified conditions, subsequent m-level suspected intrusion points are judged. If the specified conditions are met, the subsequent m-level suspected intrusion points are included in the i-th data group; otherwise, the (i+1)-th data group is created as a new data group.

[0010] As a further technical solution, the specified conditions include: The spatial distance between the suspected intrusion start coordinate of the subsequent m-level suspected intrusion point and the suspected intrusion start coordinate of the latest m-level suspected intrusion point included in the i-th data group is less than or equal to a preset graded distance threshold. The hierarchical distance threshold is set according to the requirements of the corresponding division area; The time difference between the timestamp of the subsequent m-level suspected intrusion point and the timestamp of the latest m-level suspected intrusion point included in the i-th data group is less than or equal to the preset hierarchical time difference value. The time difference for the grading is set according to the requirements of the corresponding division area.

[0011] As a further technical solution, the process of generating and judging the intrusion fitting trajectory includes: For each data group, the m-level suspected intrusion points within the data group are sorted according to the timestamp ascending order to obtain an ordered sequence of suspected intrusion points; Using the metadata of the m-level suspected intrusion points in the ordered suspected intrusion point sequence as the starting point of the trajectory, the subsequent m-level suspected intrusion points are connected sequentially according to the sequence sorting order to form a data grouped intrusion trajectory; The intrusion trajectory of the data group is judged based on the trajectory validity judgment index. If the trajectory validity judgment index is met, it is judged as a valid intrusion trajectory. The criteria for determining the validity of the trajectory include: the trajectory length is greater than or equal to the preset distance threshold of the corresponding region classification and the trajectory duration is greater than or equal to the preset duration.

[0012] As a further technical solution, the process of generating and judging the intrusion fitting trajectory includes: All data groups are sorted in chronological order based on their estimated occurrence time. The correlation is determined based on the distance between the endpoints of the starting coordinates of the (i+1)th data group and the ending coordinates of the i-th data group, and the time difference between the starting time of the (i+1)th data group and the ending time of the i-th data group. If a correlation is determined, the starting point of the (i+1)th data group is connected to the ending point of the ith data group to generate an intrusion fitting trajectory; otherwise, it is determined to be another unrelated intrusion trajectory.

[0013] As a further technical solution, the process of obtaining the intrusion classification early warning result includes: For a single actual intrusion trajectory, extract the key parameters corresponding to the actual intrusion trajectory; Based on the required alert sensitivity for the perimeter monitoring task, select an estimation algorithm with appropriate intensity sensitivity, and calculate the early warning and alert valuation based on the key parameters. The key parameters include real-time approach velocity, real-time approach acceleration, and the level parameter of the m-level division region to which the coordinates of the actual intrusion trajectory endpoint belong.

[0014] As a further technical solution, the process of generating and judging the intrusion fitting trajectory also includes: For each of the aforementioned intrusion fitting trajectories, determine whether it intersects with the boundary of the sector monitoring area; If it exists, the intersection point with the earliest timestamp is taken as the starting point of the intrusion fitting trajectory; otherwise, the intersection point of the outward extension line of the trajectory starting point based on the intrusion fitting trajectory and the boundary of the sector monitoring area is taken as the starting point of the intrusion fitting trajectory. The number of intrusion actions is obtained based on the number of initial intrusion points.

[0015] As a further technical solution, the process of forming the first monitoring surface includes: The main sensing optical cable is fixed along the outline of the perimeter carrier and the mesh laying method is adapted to the scenario. Gravity sensors are installed at both ends of the top of the carrier and the probes of the gravity sensors are placed close to the main optical cable suspension nodes; The main optical cable and the gravity sensor are respectively connected to the data acquisition device, and then the data acquisition device is connected to the back-end monitoring host; If the gravity sensor detects a gravity change amplitude greater than a preset threshold and the signal duration is greater than a preset safe duration, an advanced warning will be triggered directly.

[0016] Infrared sensing units are symmetrically arranged on both sides of the lower end of the carrier, so that the transmitting end and receiving end of each unit form a continuous infrared detection light curtain. The infrared sensing unit and the gravity sensor are simultaneously connected to the data acquisition device; If the infrared sensor detects an occlusion trigger signal with a duration greater than or equal to a preset duration and the occlusion intensity of the occlusion trigger signal is greater than or equal to a preset threshold, an advanced warning is directly triggered.

[0017] In summary, this application includes at least one of the following beneficial technical effects: 1. By extending a secondary sensor optical cable from the main sensor optical cable, a second monitoring surface that is not coplanar with the first monitoring surface is constructed, forming a three-dimensional monitoring system. This breaks through the limitations of traditional single-plane monitoring methods, enabling the capture of intrusion behavior information from different angles, reducing monitoring blind spots, achieving accurate positioning of intrusion trajectories, and improving the system's ability to capture intrusion trajectories.

[0018] 2. This invention makes scientific and effective judgments on intrusion behavior by fitting intrusion trajectories, implements false alarm prevention strategies for non-threatening behaviors, and sets different monitoring standards and early warning strategies according to the perimeter's early warning and alert intensity. This enables scientific monitoring of different areas, improves the system's ability to judge intrusion behavior, reduces false alarm rates, and avoids unnecessary time and manpower costs. Attached Figure Description

[0019] Figure 1 This is a flowchart of a perimeter intrusion detection method based on fiber optic sensing technology.

[0020] Figure 2 This is a top view of the sector-shaped area divided on the second monitoring surface.

[0021] Figure 3 This is a schematic diagram of dual-channel unidirectional netting and underground installation.

[0022] Figure 4This is a schematic diagram of a dual-channel, bidirectional optical cable radiating installation.

[0023] Figure 5 This is a schematic diagram of a dual-channel anti-shear closed-loop installation. Detailed Implementation

[0024] The embodiments of this application are described in detail below, and examples of the embodiments are shown in the accompanying drawings.

[0025] In the description of this specification, the references to "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples" refer to specific features, structures, materials, or characteristics described in connection with the described embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0026] This embodiment discloses a perimeter intrusion monitoring method based on fiber optic sensing technology. Taking the perimeter monitoring of a large industrial park as an application scenario, it employs a precise positioning vibration fiber optic system and a single-core sensing fiber optic cable as the core monitoring facilities. The vibration fiber optic system is based on Φ-OTDR technology and can simultaneously locate numerous intrusion points, displaying the monitoring results intuitively in the software interface.

[0027] Generally, other perimeter intrusion prevention devices such as video surveillance for evidence collection, pulse fences, and laser beams are also used. However, in cases where the perimeter monitoring range is large, there may be insufficient coverage or malfunction of video equipment. Therefore, this application provides a new perimeter intrusion detection method, such as... Figure 1 and Figure 3 As shown, it includes: Step S1: Along the outline of the industrial park wall, the main sensing optical cable is laid in an S-shape using a mesh-laying method. It is fixed to the guardrail bracket at the top of the wall every 1.5 meters with clips to form a closed first monitoring surface that covers the entire perimeter. Then, a secondary sensing optical cable is led out from the main sensing optical cable or the precise positioning type vibration fiber optic system. The secondary optical cable is also laid in an S-shape and extended to the outside of the wall before being fixed to the ground to form a second monitoring surface that is not coplanar with the first monitoring surface and is perpendicular to the first monitoring surface. Based on the monitoring range of the deployed optical sensing cable, such as Figure 2 As shown, the second monitoring surface is divided into multiple adjacent and equal-sized sector monitoring areas. The center of each sector monitoring area is 5m away from the outer boundary of the second monitoring surface, the sector angle is 45°, and the distance between adjacent areas is 2m, to ensure that there are no blind spots or excessive overlap in the monitoring area. By utilizing the coverage characteristics of the fan-shaped monitoring area, full coverage of any intrusion direction within the perimeter space is achieved, ensuring that when an intrusion approaches the monitoring area from any direction and any initial position outside the perimeter, its trajectory will be captured by the monitoring range of the fan-shaped monitoring area.

[0028] In this embodiment, an XYZ coordinate system is established with the center of one of the sector monitoring areas as the origin. The number of overlaps between the division areas of each sector monitoring area can be counted. For example, if a division area of ​​a sector monitoring area overlaps with an adjacent sector monitoring area once, the corresponding level is m=2. Step S2: The system uses a precise positioning type vibration fiber optic host as a detection device. The host emits a laser signal that is transmitted along a single-core sensing fiber optic cable. When there is vibration in the outside, the phase of the scattered light changes by one-thousandth and returns along the sensing fiber optic cable. After photoelectric conversion and signal analysis by the host, a suspected intrusion signal is obtained. Within a specified time period of 10 seconds, intrusion point information q((x, y, z), t) in each sector monitoring area is acquired at a frequency of 10Hz to form a set of suspected intrusion classification signals; where (x, y, z) are the location coordinates of the intrusion point and t is the timestamp; Step S3: Based on the spatial location coordinates and timestamps of the intrusion points monitored within a specified time period, group the suspected intrusion points. For example: The zero-overlap area of ​​a certain sector monitoring zone yields five intrusion point information points: a1 (10.2, 30.5, 0, 0.1), a2 (10.3, 30.2, 0, 0.2), a3 (10.4, 29.8, 0, 0.3), a4 (10.8, 29.6, 0, 0.4), and a5 (10.9, 29.3, 0, 0.5). These are collectively referred to as the first-level suspected intrusion point set A. The zero-overlap area of ​​another sector monitoring zone yielded five intrusion point information points: b1 (11.5, 30.4, 0, 0.2), b2 (11.6, 30.1, 0, 0.3), b3 (11.7, 29.7, 0, 0.4), b4 (12.1, 29.5, 0, 0.5), and b5 (12.2, 29.2, 0, 0.6), collectively referred to as the first-level suspected intrusion point set B. The first overlap of the above-mentioned sector monitoring area yields four intrusion point information points: c1 (11.2, 29.4, 0, 0.6), c2 (11.3, 29.1, 0, 0.7), c3 (11.4, 28.9, 0, 0.8), and c4 (11.5, 28.7, 0, 0.9), collectively referred to as the secondary suspected intrusion point set C; In this embodiment, the grouping criteria for the first-level suspected intrusion point set are that adjacent points with a spatial distance ≤ 0.5m and a time difference ≤ 0.2s are grouped into the same data group, and the grouping criteria for the second-level suspected intrusion point set are that the time difference ≤ 0.1s.

[0029] Thus, the set of suspected intrusion points A can be divided into the first data group A1 and the second data group A2; where the first data group A1 is {a1, a2, a3} and the second data group A2 is {a4, a5}. The set of suspected intrusion points B is divided into a first data group B1 and a second data group B2; where the first data group B1 is {b1, b2, b3} and the second data group B2 is {b4, b5}. The set of suspected secondary intrusion points C is divided into a first data group C1 and a second data group C2; where the first data group C1 is {c1, c2} and the second data group C2 is {c3, c4}. After obtaining the data groups, the intrusion fitting trajectory is generated and judged based on the data groups: For each data group, the suspected intrusion points within the data groups A1, A2, B1, B2, C1, and C2 are sorted according to the timestamp ascending order to obtain an ordered sequence of suspected intrusion points OA1, OA2, OB1, OB2, OC1, and OC2. Using the metadata of the suspected intrusion points in the ordered suspected intrusion point sequence OA1, OA2, OB1, OB2, OC1 and OC2 as the starting point of the trajectory, the subsequent suspected intrusion points are connected in sequence according to the sequence sorting order to form a data grouped intrusion trajectory; Step S4: Determine the intrusion fitting trajectory, identify and locate valid intrusion behaviors, and obtain intrusion classification alarm results; This invention determines the intrusion fitting trajectory in two ways, the first of which includes: The intrusion trajectory of the data group is judged based on the trajectory validity judgment index. If the trajectory validity judgment index is met, it is judged as a valid intrusion trajectory. The criteria for determining the validity of the trajectory include: The trajectory length in the first-level suspected intrusion point set A is greater than or equal to 0.5m, and the trajectory duration is greater than or equal to 0.2s; therefore, OA1 and OA2 are both valid intrusion trajectories. The trajectory lengths in the first-level suspected intrusion point set B are greater than or equal to 0.5m, and the trajectory durations are greater than or equal to 0.3s; therefore, OB1 and OB2 are both valid intrusion trajectories. The trajectory length in the secondary suspected intrusion point set C is greater than or equal to 0.2m, and the trajectory duration is greater than or equal to 0.2s; therefore, OC1 and OC2 are both valid intrusion trajectories. All data groups are sorted in chronological order based on their estimated occurrence time. The estimated occurrence time is the center time point of the data group, and the method for calculating the estimated occurrence time includes:

[0030] in, Values ​​at the center time point, For the data grouping Timestamps of each intrusion point This represents the total number of intrusion points in the data group; The calculated center time values ​​are 0.2 for A1, 0.3 for A2, 0.3 for B1, 0.36 for B2, 0.43 for C1, and 0.57 for C2. All data groups are sorted in chronological order based on their estimated occurrence time. The correlation is determined based on the distance between the endpoints of the starting coordinates of the (i+1)th data group and the ending coordinates of the i-th data group, and the time difference between the starting time of the (i+1)th data group and the ending time of the i-th data group. If a correlation is determined, the starting point of the (i+1)th data group is connected to the ending point of the ith data group to generate an intrusion fitting trajectory; otherwise, it is determined to be another unrelated intrusion trajectory.

[0031] Calculate the endpoint distance and time connection difference between the intrusion fitting trajectories of the adjacent partitioned regions; The method for calculating the endpoint distance of the intrusion fitting trajectory includes:

[0032] in, The spatial distance between two points. For point of Axis coordinates For point of Axis coordinates For point of Axis coordinates For point of Axis coordinates For point of Axis coordinates For point of Axis coordinates; Calculate the endpoint distances of C1 and C2 to A1, A2, B1, and B2 respectively to obtain two intrusion fitting trajectories L1 and L2; L1: OA1→OA2→OC1; L2: OB1→OB2→OC2; The second scenario includes: constructing an AI recognition model based on machine learning algorithms for the intrusion fitting trajectory; The AI ​​recognition model was pre-trained with samples, and the training sample set included standard trajectory features and associated vibration signal features corresponding to human intrusion, human sabotage, large animal sabotage, and natural disaster intrusion. The core feature parameters of the intrusion fitting trajectory are input into the trained AI recognition model and matched with the standard features in the sample set. The core feature parameters include the real-time approach speed of the intrusion fitting trajectory, the real-time approach acceleration of the intrusion fitting trajectory, and the dwell time. Based on the matching results of the AI ​​recognition model, the specific type of the intrusion fitting trajectory is output; When the matching similarity is higher than the preset threshold of 65%, it is determined to be an intrusion of the corresponding type of alarm behavior; when the matching similarity is lower than the preset threshold of 30%, it is determined to be an intrusion of non-alarm behavior. If the non-alarm intrusion behavior causes damage to the perimeter, the core feature parameters of the non-alarm intrusion behavior are recorded in the AI ​​recognition model.

[0033] In this embodiment, the AI ​​recognition model further includes an intrusion intent prediction module; An intrusion intent classification model was constructed and trained using a random forest algorithm. The input layer of the intrusion intent classification model was set as the core feature parameters of the intrusion fitting trajectory, including trajectory coverage, movement speed and acceleration, dwell time, trajectory repetition count, and distance to perimeter protection facilities. The output layer of the intrusion intent classification model was the probability value of three types of intrusion intent: probing intrusion, forced breakthrough intrusion, and accidental intrusion. The core feature parameters of the extracted intrusion fitting trajectory are input into the intrusion intent classification model, and the target intrusion intent probability of the intrusion fitting trajectory is output. When the probability of the target intrusion intent is higher than a preset high intent threshold of 70%, the target intrusion intent is determined to be a forced breakthrough intrusion, and an alarm mechanism is triggered. When the probability of the target intrusion intent is less than 70% of the preset high intent threshold and greater than 30% of the preset low intent threshold, the target intrusion intent is determined to be a probing intrusion and is marked as a key intrusion.

[0034] When the probability of the target intrusion intent is lower than the preset low intent threshold of 30%, the target intrusion intent is determined to be a false intrusion, and no alarm is triggered.

[0035] In this embodiment, the AI ​​recognition model further includes an environmental feature learning module; Real-time collection of environmental parameters such as temperature, humidity, and wind speed; Beforehand, environmental data under different environmental conditions are collected through experiments. The background vibration signal characteristics of the sensing optical cable and the vibration signal characteristics corresponding to typical intrusion behaviors are recorded simultaneously under each environmental condition to establish an environmental signal feature mapping database. During the monitoring process, the environmental signal feature mapping database is called in real time, and the corresponding signal feature data is matched according to the currently collected environmental parameters to dynamically adjust the vibration signal acquisition parameters to counteract environmental noise interference. Adjusting the feature matching weights of the AI ​​recognition model allows it to adapt to signal variation patterns under different environments, thereby improving the accuracy of intrusion type identification.

[0036] In this embodiment, the method for locating effective intrusion behavior includes: For the intrusion behavior detected by the AI ​​recognition algorithm, a spatial coordinate mapping model is constructed by combining the map of the industrial park and the three-dimensional monitoring system composed of the main sensor optical cable and the slave sensor optical cable. The system calculates and outputs the specific location information of the intrusion caused by the alarm behavior, locates the intrusion, and transmits the information to the terminal.

[0037] In this embodiment, the process of obtaining the intrusion classification early warning result includes: For a single intrusion fitting trajectory, extract the key parameters corresponding to the intrusion fitting trajectory; Based on the required alert sensitivity for the perimeter monitoring task, select an estimation algorithm with appropriate intensity sensitivity, and calculate the early warning and alert valuation based on the key parameters. The key parameters include the real-time approach velocity of the intrusion fitting trajectory, the real-time approach acceleration of the intrusion fitting trajectory, and the level parameter of the m-level division region to which the endpoint coordinates of the actual intrusion trajectory belong.

[0038] In the case of a perimeter monitoring scenario with high alert level, a high-sensitivity calculation method is adopted so that when the intrusion behavior reaches the early area of ​​the second monitoring surface, the early warning and alert value will show a significant upward trend, thus achieving early high-intensity early warning.

[0039] in, For high-sensitivity early warning and alert valuation, This is the basic warning coefficient for the perimeter. The region level where the final location of the fitted invasion trajectory is located. , The real-time approach speed of the intrusion fitting trajectory. To fit the intrusion trajectory in real time to the acceleration, The coefficient representing the actual approach velocity of the intrusion fitting trajectory; The real-time approach velocity of the intrusion trajectory L1 is v = 3.2 m / s, and the real-time approach acceleration of the intrusion trajectory is a = 0.8 m / s². 2 The endpoint is located in the secondary region k=2. =10; the high-sensitivity early warning and alert value is calculated. =1.529; If the aforementioned warning estimate is less than or equal to 2.5, a primary warning is triggered; If the warning estimate is greater than 2.5, a high-level warning is triggered; therefore, L1 triggers a primary warning at this time. For perimeter monitoring scenarios with low alert levels, a low-sensitivity calculation method is used.

[0040] in, For low-sensitivity early warning and alert valuation, The coefficient of acceleration is used to fit the trajectory of a low-sensitivity intrusion in real time. Taking the intrusion trajectory L2 as an example, the real-time approach velocity of the intrusion trajectory L2 is extracted as v = 3.5 m / s, and the real-time approach acceleration of the intrusion trajectory L2 is extracted as a = 0.6 m / s². 2 The endpoint of the intrusion fitting trajectory L2 is located in the secondary region k=2. =0.1; Calculated ; The low-sensitivity calculation method ensures that when intrusion occurs in the early stage of the second monitoring surface, the warning and alert value increases gradually without triggering a warning action. However, once the intrusion enters the later stage of the second monitoring surface, the warning and alert value increases rapidly and significantly, reaching a level greater than or equal to the preset warning value. When this happens, a high-level early warning is triggered directly to prevent false alarms from earlier warnings.

[0041] In this embodiment, the process of generating and judging the intrusion fitting trajectory further includes: For the intrusion fitting trajectory L1 and the intrusion fitting trajectory L2, it is determined that the intrusion fitting trajectory L1 and the intrusion fitting trajectory L2 do not intersect with the boundary of the sector monitoring area; Therefore, the number of intersections with the boundary of the sector monitoring area is obtained by extending the trajectory from the starting point of the intrusion fitting trajectory L1 and the intrusion fitting trajectory L2 outwards. At this time, the number of intersections is 2, and the number of intrusion behaviors is 2.

[0042] The present invention also addresses the following specific requirements for certain scenarios: Gravity sensors are installed at both ends of the top of the carrier and the probes of the gravity sensors are placed close to the main optical cable suspension nodes; The main optical cable and the gravity sensor are respectively connected to the data acquisition device, and then the data acquisition device is connected to the back-end monitoring host; If the gravity sensor detects a change in gravity signal, and the gravity is greater than the safe gravity threshold of 2N and the duration is greater than the safe duration of 2 seconds, the data acquisition device will immediately transmit the signal to the back-end host and directly trigger an advanced warning. Infrared sensing units are symmetrically arranged on both sides of the lower end of the carrier, so that the transmitting end and receiving end of each unit form a continuous infrared detection light curtain. The infrared sensing unit and the gravity sensor are simultaneously connected to the data acquisition device; If the infrared sensor unit detects that the occlusion lasts for more than or equal to a preset duration of 1 second and the occlusion intensity is greater than or equal to a preset threshold of 50%, the infrared sensor unit will trigger a signal to directly trigger a level one warning.

[0043] Under extreme weather conditions, false vibration signals caused by environmental noise are eliminated through multi-dimensional verification using gravity sensors, infrared sensing units, and fiber optic sensing data. The vibration signal of a set of suspected intrusion points was verified by the gravity sensor to show no change in gravity, and the infrared sensor showed no obstruction. It was determined to be environmental interference and was not included in the set of suspected intrusion classification signals, which effectively reduced the false alarm rate. When a real intrusion occurs, the fiber optic sensor captures the continuous vibration trajectory, the gravity sensor detects the change in gravity, and the infrared sensor triggers an obstruction signal. When these three data points are consistent, the system is accurately identified as a valid intrusion, ensuring stable operation of the system even in harsh environments.

[0044] This embodiment, through actual deployment and testing, verifies that the perimeter intrusion monitoring method based on fiber optic sensing technology can accurately locate, track, and provide graded early warnings for perimeter intrusion behavior in industrial parks, with a false alarm rate of less than 3% and a response time of less than or equal to 2 seconds, meeting the actual needs of perimeter security monitoring.

[0045] When the alarm is triggered, the surveillance camera captures real-time video footage of the intrusion area. After confirming the intrusion, the sound and light alarm emits a high-decibel sound and bright light to deter the intruder and transmit the alarm location information to the on-site security personnel.

[0046] In this embodiment, the deployment method of the monitoring system may also include dual-channel unidirectional grid connection and underground installation, such as... Figure 3 As shown: The fiber optic cable lines laid in the net are designated as Channel 1, serving as the first line of defense, while the fiber optic cable lines laid underground are designated as Channel 2, serving as the second line of defense. The channel 1 and the channel 2 can be set up with independent perimeter defense zones. When the optical cable of one channel is damaged, the optical cable of the other channel can still operate normally, ensuring the continuous protection of the entire perimeter defense zone by the system.

[0047] In this embodiment, the installation method may further include dual-channel bidirectional optical cable radiating installation, such as... Figure 4 As shown: Channel 1 adopts a wave-shaped mesh laying method, with a 50cm spacing between the optical cable and the wire mesh fixing points, and the wave amplitude controlled between 30-50cm to ensure vibration transmission sensitivity. Channel 2 uses a wire mesh laid in a straight line with a fixed point spacing of 1 meter, and is laid parallel to Channel 1 with a spacing of 1.5 meters. With the dual-channel bidirectional laying design, the detection perimeter of the system can be extended to 80KM based on the maximum detection distance of 40KM for a single channel of the precision positioning vibration fiber optic host, thus adapting to the protection needs of longer perimeters. Channel 1 and Channel 2 can be monitored independently, improving the overall monitoring coverage and reliability.

[0048] In this embodiment, the installation method may further include dual-channel anti-shear closed-loop installation, such as... Figure 5 As shown: The optical cable of channel 1 is laid clockwise from the beginning to the end along the perimeter defense zone, and the optical cable of channel 2 is laid counterclockwise from the end to the beginning along the perimeter defense zone, forming a double optical cable ring network structure. When the optical cables of Channel 1 and Channel 2 at a certain point in the perimeter defense zone are damaged at the same time, the system can still achieve signal transmission and monitoring through the remaining intact segments of the ring network, ensuring continuous protection of the entire perimeter defense zone and avoiding the interruption of overall protection due to local optical cable damage.

[0049] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A perimeter intrusion detection method based on fiber optic sensing technology, characterized in that, Includes the following steps: Step S1: Deploy the main sensing optical cable according to the perimeter contour to form a first monitoring surface, and draw out the secondary sensing optical cable from the main sensing optical cable to form a second monitoring surface that is not coplanar with the first monitoring surface; Step S2: Process the vibration signal of the sensing optical cable and obtain a set of suspected intrusion classification signals within a specified time period at equal intervals. Step S3: Based on the core data of the suspected intrusion classification signal set, suspected intrusion points are grouped and intrusion fitting trajectories are generated; Step S4: Determine the intrusion fitting trajectory, identify and locate valid intrusion behaviors, and obtain intrusion classification alarm results.

2. The perimeter intrusion monitoring method according to claim 1, characterized in that, The method for processing the vibration signal of the sensing optical cable includes: The vibration fiber optic host emits a laser signal along the sensing optical cable. When there is external vibration, the sensing optical cable converts the vibration signal into scattered light and transmits it back to the vibration fiber optic host. The vibration fiber optic host sequentially performs photoelectric conversion processing on the returned scattered light and collects and analyzes the vibration signal to finally obtain the suspected intrusion signal.

3. The perimeter intrusion detection method based on fiber optic sensing technology according to claim 1, characterized in that, The method for obtaining the suspected intrusion classification signal set includes: The second monitoring surface is divided into multiple sector-shaped monitoring areas, and key parameters of the sector-shaped monitoring areas are configured, including the center position, sector angle, and interval distance. The level m of the corresponding division region is determined based on the number of overlaps of each division region in the sector-shaped monitoring area; For each sector-shaped monitoring area, the set of suspected intrusion signals obtained by the vibration fiber optic host is taken as the m-level suspected intrusion point set; the set of suspected intrusion classification signals is the set of the m-level suspected intrusion point set.

4. The perimeter intrusion monitoring method based on fiber optic sensing technology according to claim 3, characterized in that, The method for grouping suspected intrusion points includes: For each set of m-level suspected intrusion points, the corresponding m-level suspected intrusion point metadata is confirmed; the suspected intrusion point metadata includes the suspected intrusion start coordinates and timestamp; Establish the i-th data group containing the metadata of the m-level suspected intrusion points; Based on specified conditions, subsequent m-level suspected intrusion points are judged. If the specified conditions are met, the subsequent m-level suspected intrusion points are included in the i-th data group; otherwise, the (i+1)-th data group is created as a new data group.

5. A perimeter intrusion detection method based on fiber optic sensing technology according to claim 4, characterized in that, The specified conditions include: The spatial distance between the suspected intrusion starting coordinates of the subsequent m-level suspected intrusion points and the suspected intrusion starting coordinates of the latest m-level suspected intrusion point included in the i-th data group is less than or equal to a preset graded distance threshold. The hierarchical distance threshold is set according to the requirements of the corresponding division area; The time difference between the timestamp of the subsequent m-level suspected intrusion point and the timestamp of the latest m-level suspected intrusion point included in the i-th data group is less than or equal to the preset hierarchical time difference value. The time difference for the grading is set according to the requirements of the corresponding division area.

6. The perimeter intrusion detection method based on fiber optic sensing technology according to claim 4, characterized in that, The process of generating and judging the intrusion fitting trajectory includes: For each data group, the m-level suspected intrusion points within the data group are sorted according to the timestamp ascending order to obtain an ordered sequence of suspected intrusion points; Using the metadata of the m-level suspected intrusion points in the ordered suspected intrusion point sequence as the starting point of the trajectory, the subsequent m-level suspected intrusion points are connected sequentially according to the sequence sorting order to form a data grouped intrusion trajectory; The intrusion trajectory of the data group is judged based on the trajectory validity judgment index. If the trajectory validity judgment index is met, it is judged as a valid intrusion trajectory. The criteria for determining the validity of the trajectory include: the trajectory length is greater than or equal to the preset distance threshold of the corresponding region classification and the trajectory duration is greater than or equal to the preset duration.

7. The perimeter intrusion detection method based on fiber optic sensing technology according to claim 4, characterized in that, The process of generating and judging the intrusion fitting trajectory includes: All data groups are sorted in chronological order based on their estimated occurrence time. The correlation is determined based on the distance between the endpoints of the starting coordinates of the (i+1)th data group and the ending coordinates of the i-th data group, and the time difference between the starting time of the (i+1)th data group and the ending time of the i-th data group. If a correlation is determined, the starting point of the (i+1)th data group is connected to the ending point of the ith data group to generate an intrusion fitting trajectory; otherwise, it is determined to be another unrelated intrusion trajectory.

8. The perimeter intrusion detection method based on fiber optic sensing technology according to claim 7, characterized in that, The process of obtaining intrusion classification early warning results includes: For a single actual intrusion trajectory, extract the key parameters corresponding to the actual intrusion trajectory; Based on the required alert sensitivity for the perimeter monitoring task, select an estimation algorithm with appropriate intensity sensitivity, and calculate the early warning and alert valuation based on the key parameters. The key parameters include real-time approach velocity, real-time approach acceleration, and the level parameter of the m-level division region to which the coordinates of the actual intrusion trajectory endpoint belong.

9. A perimeter intrusion detection method based on fiber optic sensing technology according to claim 7, characterized in that, The process of generating and judging the intrusion fitting trajectory also includes: For each of the aforementioned intrusion fitting trajectories, determine whether it intersects with the boundary of the sector monitoring area; If it exists, the intersection point with the earliest timestamp is taken as the starting point of the intrusion fitting trajectory; otherwise, the intersection point of the outward extension line of the trajectory starting point based on the intrusion fitting trajectory and the boundary of the sector monitoring area is taken as the starting point of the intrusion fitting trajectory. The number of intrusion actions is obtained based on the number of initial intrusion points.

10. A perimeter intrusion detection method based on fiber optic sensing technology according to claim 1, characterized in that, The process of forming the first monitoring surface includes: The main sensing optical cable is fixed along the outline of the perimeter carrier and the mesh laying method is adapted to the scenario. Gravity sensors are installed at both ends of the top of the carrier and the probes of the gravity sensors are placed close to the main optical cable suspension nodes; The main optical cable and the gravity sensor are respectively connected to the data acquisition device, and then the data acquisition device is connected to the back-end monitoring host; If the gravity sensor detects a gravity change amplitude greater than a preset threshold and the signal duration is greater than a preset safe duration, an advanced warning is triggered directly. Infrared sensing units are symmetrically arranged on both sides of the lower end of the carrier, so that the transmitting end and receiving end of each unit form a continuous infrared detection light curtain. The infrared sensing unit and the gravity sensor are simultaneously connected to the data acquisition device; If the infrared sensor detects an occlusion trigger signal with a duration greater than or equal to a preset duration and the occlusion intensity of the occlusion trigger signal is greater than or equal to a preset threshold, an advanced warning is directly triggered.

Images