Intelligent optical fiber perimeter alarm device and method with adjustable vibration sensitivity

Abstract

The invention discloses an intelligent optical fiber perimeter alarm device and method with adjustable vibration sensitivity, the device comprises a light source, a beam splitter and an optical fiber link connected with the beam splitter through an optical fiber for communication, and the optical fiber link comprises a 2 * 2 coupler, an optical fiber for time delay and a 1 * 2 coupler; and each group of light links also comprises a reflecting mirror connected with the 1 * 2 coupler through a sensing optical fiber and a detector connected with the 2 * 2 coupler. According to the intelligent optical fiber perimeter alarm method, the sensitivity of the vibration signal on a hardware platform can be adjusted, the intelligent optical fiber perimeter alarm method has the vibration signal recognition capability required by the intelligent low hardware platform, and the vibration sensitivity is adjustable, after the vibration signal is collected, eight feature values such as a time domain, a frequency domain and a Mel coefficient are extracted from the collected signal to form a feature vector, and the feature vector is extracted; a support vector machine (SVM) is used for judgment, feature values are extracted from different types of vibration signals, a vibration signal mode library is formed, and the conditions of raining, handrail shaking, no disturbance and the like are effectively recognized and distinguished.

Description

Technical Field

[0001] The present invention relates to the technical field of alarm systems, and in particular to an intelligent optical fiber perimeter alarm device with adjustable vibration sensitivity and a method thereof. Background Art

[0002] Perimeter security, also known as intrusion detection or fence detection, refers to the prevention of the boundaries of a specific area to prevent illegal intrusion of objects outside the area or the escape of people in the area. The perimeter security system monitors the protected area. When an intrusion occurs, the detector will immediately send out an alarm signal and notify the user through the relevant alarm linkage equipment, thereby achieving effective prevention, such as infrared counter-radiation, pulse electronic fence, vibration cable, video surveillance, etc. These measures are applied to different places according to their different advantages, but there are also certain defects. For example, infrared counter-radiation requires the direct installation of detectors, which is greatly affected by the terrain and is only used in scenes with relatively open areas and small defense areas; pulse electronic fences consume a lot of power and there is a risk of generating electric sparks, so they cannot be used in flammable and explosive places; buried cable detection is not suitable for large-scale laying, and the monitoring area is limited.

[0003] The fiber optic perimeter alarm system uses optical fiber to sense external vibration signals. Compared with other types of perimeter alarms, it has the characteristics of being passive, intrinsically safe, long-distance, and 24-hour uninterrupted monitoring. It is increasingly widely used in the field of perimeter security. However, in the existing patents on fiber optic perimeter alarm systems, the vibration sensitivity of the fiber optic perimeter alarm system is difficult to adjust through hardware, and is significantly disturbed by environmental vibration, which has a great impact on intrusion monitoring; in addition, the existing fiber optic perimeter alarm system lacks intelligent vibration signal recognition capabilities, or the algorithms used are complex and have high hardware requirements, making it difficult for the fiber optic perimeter alarm system to be widely used.

[0004] For example, patent CN101871809B discloses a zone-type optical fiber distributed vibration sensor and a perimeter alarm system using the sensor. The system uses a circular Sagnac interferometer system. Although the polarization fading phenomenon in the optical fiber is avoided by devices such as depolarizers, the sensitivity of the system to vibration signals is difficult to adjust through hardware, and the system lacks the ability to recognize vibration signals.

[0005] Patent CN110706435A discloses a fiber perimeter intrusion alarm system based on a 3*3 coupler. The system uses a DFB laser and a 3*3 coupler to form a ring interference structure, which can reduce the impact of polarization fading on system performance. However, the system is difficult to adjust the sensitivity of vibration signals through hardware, and lacks intelligent vibration signal recognition capabilities.

[0006] Patent CN111597994A discloses a fiber perimeter security intrusion event recognition model construction method and security system. The system uses multimode optical fiber as a sensing unit and uses singular values ​​and kurtosis values ​​to form vectors for vibration signal recognition. However, the sensitivity of the system to vibration signals is difficult to adjust through hardware, and the use of complex neural network models requires a high-performance hardware platform, which significantly increases the system cost and is not conducive to the widespread application of fiber perimeter systems. Summary of the invention

[0007] In order to solve the technical problems mentioned in the background technology, the present invention provides an intelligent optical fiber perimeter alarm device and method with adjustable vibration sensitivity.

[0008] The present invention adopts the following technical scheme: an intelligent optical fiber perimeter alarm device with adjustable vibration sensitivity, comprising a signal processing system, a light source, a beam splitter, and a plurality of parallel optical links connected to the beam splitter through a communication optical fiber; each group of optical links comprises a *coupler, a delay optical fiber, and a *coupler connected in sequence through a communication optical fiber; each group of optical links also comprises a reflector connected to the *coupler through a sensing optical fiber and a detector connected to the *coupler; the output light of the light source is divided into multiple beams by the beam splitter, and each beam of light is divided into two paths of light by the *coupler after passing through the communication optical fiber, one of which enters the delay optical fiber and is connected to the *coupler; the other path is directly connected to another input port of the *coupler; the two paths of light are coupled into a beam of light by the *coupler and then enter the sensing optical fiber, and are reflected back to the *coupler at the reflector to form interference light, and the interference light is received by the photoelectric detector, and the detector performs photoelectric conversion on the received interference light and transmits it to the signal processing system to extract signal characteristic values, and identifies it to determine whether to issue an alarm signal.

[0009] Furthermore, the wavelength of the light source ranges from visible light to infrared light, preferably 1310 to 1650 nm.

[0010] Furthermore, the number of channels of the beam splitter may be 1-256.

[0011] Furthermore, the length of the communication optical fiber ranges from 0.1m to 100km, the length of the delay optical fiber 4 ranges from 1m to 20km, and the length of the sensing optical fiber 6 ranges from 0.1m to 200km.

[0012] Furthermore, the beam splitter, 2*2 coupler, and 1*2 coupler adopt equal-ratio beam splitting or non-equal-ratio beam splitting.

[0013] Furthermore, the reflectivity of the reflector ranges from 1% to 100%.

[0014] The present invention also provides an intelligent optical fiber perimeter alarm method with adjustable vibration sensitivity of an intelligent optical fiber perimeter alarm device with adjustable vibration sensitivity, comprising the following steps:

[0015] S1: Establish a vibration signal recognition model, including:

[0016] S101: The beam splitter channel divides the light output from the light source into two paths through the 2*2 coupler, one of which enters the delay optical fiber and is connected to the 1*2 coupler; the other path is connected to the other input port of the 1*2 coupler; the two paths of light are coupled into one beam through the 1*2 coupler and then enter the sensing optical fiber, and are reflected back to the 2*2 coupler at the reflector to form interference light, which is received by the photodetector;

[0017] S102: The detector transmits the received signal to the signal processing system, and the signal processing system extracts the effective vibration signal, and extracts eigenvalue 1 and eigenvalue 2 according to the zero-crossing energy and the vibration signal ratio; extracts eigenvalue 3 and eigenvalue 4 according to the frequency domain signal band variance and spectral entropy; calculates the Mel coefficient and extracts eigenvalues ​​5 to 8;

[0018] S103: Combining the above eigenvalues ​​into an eigenvalue vector;

[0019] S104: Collecting multiple groups of eigenvalue vectors of different types of vibration signals to form a vibration signal recognition model database;

[0020] S105: Bringing the vibration signal recognition model database into the SVM for training to generate a vibration signal recognition model;

[0021] S2: Using the generated vibration signal recognition model to recognize the vibration signal collected in real time through the above steps, and judging the intrusion behavior, including:

[0022] S201: adopting the above steps S101-S103, collecting the eigenvalues ​​of the vibration signal in real time, and forming a real-time signal eigenvalue vector;

[0023] S202: Bring the above-mentioned real-time signal eigenvalue vector into SVM, use the vibration signal recognition model to identify the signal, and classify it according to the similarity matching between the eigenvector of the real-time collected signal and the eigenvector of each type of vibration data in the vibration signal recognition model to determine the intrusion behavior.

[0024] Compared with the prior art, the advantages of the present invention are: the intelligent fiber optic perimeter alarm device with adjustable vibration sensitivity designed by the present invention can adjust the sensitivity of vibration signals on the hardware platform, has the intelligent vibration signal recognition capability with low hardware platform requirements, is safe and reliable, and has the advantages of 24-hour uninterrupted intelligent monitoring. The intelligent fiber optic perimeter alarm method with adjustable vibration sensitivity, after collecting vibration signals through the intelligent fiber optic perimeter alarm device with adjustable vibration sensitivity, extracts 8 eigenvalues ​​such as time domain, frequency domain and Mel coefficient from the collected signals to form a eigenvector, and uses the support vector machine SVM for discrimination. By extracting eigenvalues ​​from a variety of different types of vibration signals, a vibration signal pattern library is formed, which can effectively identify and distinguish between rain, shaking railings, no disturbance, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is a structural schematic diagram of an intelligent optical fiber perimeter alarm device with adjustable vibration sensitivity in an embodiment of the present invention;

[0026] Figure 2 is a vibration signal collected under different delayed optical fiber lengths in an embodiment of the present invention;

[0027] Figure 3 is a description of characteristic values ​​in an embodiment of the present invention;

[0028] Figure 4 It is a schematic diagram of the flow of an intelligent optical fiber perimeter alarm method with adjustable vibration sensitivity in an embodiment of the present invention;

[0029] FIG5 is a diagram of multiple groups of original signals for three types of events, wherein FIG5(a) is an original signal diagram for an undisturbed event, FIG5(b) is an original signal diagram for a raining event, and FIG5(c) is an original signal diagram for a railing shaking event;

[0030] Figure 6 It is a schematic diagram of the result of signal classification using feature quantity 3 and feature quantity 4 for the three types of events in FIG5 .

[0031] in:

[0032] 1-light source, 2-beam splitter, 3-2*2 coupler, 4-optical fiber for delay, 5-1*2 coupler, 6-optical fiber for sensing, 7-reflector, 8-detector, 9-optical fiber for communication. DETAILED DESCRIPTION

[0033] In order to facilitate those skilled in the art to understand the technical solution of the present invention, further description will be given below with reference to the accompanying drawings. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention.

[0034] In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of embodiments of the present invention. However, it is apparent that one or more embodiments may be implemented without these specific details. In addition, in the following description, descriptions of known structures and techniques are omitted to avoid unnecessary confusion of concepts of the present invention.

[0035] The invention relates to an intelligent optical fiber perimeter alarm device with adjustable vibration sensitivity, comprising a signal processing system, a light source 1, a beam splitter 2, and multiple groups of parallel optical links connected to the beam splitter 2 via a communication optical fiber 9; each group of optical links comprises a 2*2 coupler 3, a delay optical fiber 4, and a 1*2 coupler 5 connected in sequence via the communication optical fiber 9; each group of optical links also comprises a reflector 7 connected to the 1*2 coupler 5 via a sensing optical fiber 6, and a detector 8 connected to the 2*2 coupler 3. Specifically, as Figure 1 As shown, it is a schematic diagram of the structure of an intelligent fiber optic perimeter alarm device with adjustable vibration sensitivity in an embodiment of the present invention. The intelligent fiber optic perimeter alarm device with adjustable vibration sensitivity includes: a light source 1, a beam splitter 2, a 2*2 coupler 3-1, a delay optical fiber 4-1, a 1*2 coupler 5-1, a sensing optical fiber 6-1, a reflector 7-1, a detector 8-1, and a communication optical fiber 9-1; a 2*2 coupler 3-2, a delay optical fiber 4-2, a 1*2 coupler 5-2, a sensing optical fiber 6-2, a reflector 7-2, a detector 8-2, and a communication optical fiber 9-2; ...; a 2*2 coupler 3-n, a delay optical fiber 4-n, a 1*2 coupler 5-n, a sensing optical fiber 6-n, a reflector 7-n, a detector 8-n, and a communication optical fiber 9-n.

[0036] As shown in the first optical link in the figure, the output light of light source 1 is divided into multiple beams by beam splitter 2. After each beam passes through communication optical fiber 9-1, it is divided into two beams by 2*2 coupler 3-1. One beam enters delay optical fiber 4-1 and connects to 1*2 coupler 5-1, and the other beam is connected to the other input port of 1*2 coupler 5-1. The two beams are coupled into one beam by 1*2 coupler 5-1, enter sensor optical fiber 6-1, and are reflected back to 2*2 coupler 3-1 at reflector 7-1 to form interference light.

[0037] When external vibration is transmitted to the sensing optical fiber 6-1, the refractive index and length of the sensing optical fiber 6-1 are adjusted, thereby affecting the phase of the light transmitted in the sensing optical fiber 6-1, and further affecting the interference signal. When there is no external vibration signal, the two paths of light forming the interference have the same optical path, so the signal has no fluctuations; when there is an external vibration signal, the two paths of light forming the interference have different times passing through the vibration point, so there is a phase difference when they meet, and the signal has fluctuations. After the detector performs photoelectric conversion on the interference light, the data processing system extracts the signal characteristic value and identifies it.

[0038] In this embodiment, the wavelength of the laser ranges from visible light to infrared light, preferably 1310 to 1650 nm.

[0039] In this embodiment, the number of channels of the beam splitter 2 can be 1 to 256.

[0040] In such Figure 1 In the intelligent optical fiber perimeter alarm device with adjustable vibration sensitivity shown in the figure, since the light does not pass through the 2*2 coupler 3-1 for beam splitting when passing through the communication optical fiber 9-1, the communication optical fiber 9-1 is insensitive to vibration. The communication optical fiber can be used to guide the optical signal from the monitoring room to the area to be monitored, which can avoid the influence of vibration. When the light passes through the link such as the coupler, optical fiber, and reflector, interference is formed at the 2*2 coupler 3-1, and the phase of the interference light can be expressed by formula (1):

[0041]

[0042] In the formula is the amplitude of the external vibration signal, and ω is the frequency of the external vibration signal.

[0043] After combining the delay optical fiber 4-1 and the sensing optical fiber 6-1 and the length of the link, the phase of the interference light can be further expressed by the following formula (2):

[0044]

[0045] Wherein Ld is the length of the delay optical fiber 4-1, L0 is the length of the sensing optical fiber 6-1, L is the distance from the disturbance point to the reflector 7-1, n is the refractive index of the optical fiber core, and c is the speed of light in a vacuum.

[0046] It can be seen from formula (2) that by changing the length of the delay optical fiber 4-1Ld, the amplitude of the phase disturbance can be adjusted. Figure 2 The lengths of the delay optical fiber 4-1 are 0km, 1km, 3km, and 5km respectively. Figure 2It can be seen that the signal amplitude detected by the system is related to the length of the delay optical fiber 4-1. By adjusting the length of the delay optical fiber 4-1, the vibration signal amplitude can be conveniently adjusted within a wide range. By adjusting the vibration signal sensitivity, the noise signal generated by weak environmental vibration can be effectively suppressed, which is conducive to improving the signal-to-noise ratio and enhancing monitoring accuracy.

[0047] In the specific implementation of this embodiment, the length range of the communication optical fiber 9-1 is 0.1m to 100km, the length range of the delay optical fiber 4-1 is 1m to 20km, and the length range of the sensing optical fiber 6-1 is 0.1m to 200km. The beam splitter 2 can use equal-ratio beam splitting or unequal-ratio beam splitting. The 2*2 coupler 3-1 can use equal-ratio beam coupling or unequal-ratio beam splitting. The 1*2 coupler 5-1 can use equal-ratio beam coupling or unequal-ratio beam splitting.

[0048] In this embodiment, the reflectivity of the reflector 7 - 1 ranges from 1% to 100%.

[0049] After extracting the vibration signal, the eigenvalues ​​of the signal are extracted, which are eigenvalue 1-zero-crossing energy, eigenvalue 2-large signal ratio, eigenvalue 3-band variance, eigenvalue 4-spectral entropy, and eigenvalues ​​5 to 8-Mel coefficients, such as Figure 3 shown.

[0050] Eigenvalue 1-zero-crossing energy, select the non-vibration signal segment as the reference signal, and judge the zero-crossing energy of the vibration signal segment, which can be expressed by formula (3):

[0051]

[0052] Where E(n) is the zero-crossing energy and T1 is the zero-crossing energy threshold.

[0053] Eigenvalue 2 - Large signal ratio, set the large signal amplitude benchmark, calculate the proportion of vibration signal segments exceeding the amplitude benchmark, which can be expressed as follows:

[0054]

[0055] Where L is the signal length and T2 is the large signal threshold setting value.

[0056] Eigenvalue 3-band variance, the average value of the deviation between each point of the vibration signal and the variance, reflects the degree of signal fluctuation, and can be expressed by formula (4):

[0057]

[0058] Where X is the signal spectrum, and S is the average value of the signal spectrum amplitude.

[0059] Eigenvalue 4-spectral entropy reflects the disorder of the signal amplitude distribution in the frequency domain and can be expressed by formula (5):

[0060] V4=-∑p(h)·logp(k)

[0061] Where p(k) represents the normalized spectral probability density of the frequency component.

[0062] Eigenvalue 5 to 8 - Mel coefficient, which reduces the impact of high-frequency changes and enhances the impact of low-frequency signals. It can be expressed by formula (6):

[0063]

[0064] Where m is the mth order Mel filter bank, n represents the spectrum line, and S(m) is the product of the spectrum line energy and the filter response function.

[0065] The characteristic signal is combined into a characteristic value vector and brought into the support vector machine (SVM) for identification, and the vibration signal type is intelligently identified. The signal processing flow chart is as follows: Figure 4 shown.

[0066] Based on the above-mentioned intelligent optical fiber perimeter alarm device with adjustable vibration sensitivity, the present invention also provides an intelligent optical fiber perimeter alarm method with adjustable vibration sensitivity, firstly establishing a signal recognition model database, which includes the following steps:

[0067] Step 1: The output light of channel 1 of beam splitter 2 is divided into two paths through 2*2 coupler 3-1, one of which enters the delay optical fiber 4-1 and is connected to 1*2 coupler 5-1; the other is connected to the other input port of 1*2 coupler 5-1. The two paths of light are coupled into one beam through 1*2 coupler 5-1 and enter the sensing optical fiber 6-1, and are reflected back to 2*2 coupler 3-1 at reflector 7-1 to form interference light, which is received by photodetector 8-1. The output light of channel 2 of beam splitter 2 is divided into two paths of light through 2*2 coupler 3-2, one of which enters the delay optical fiber 4-2 and is connected to 1*2 coupler 5-2; the other is connected to the other input port of 1*2 coupler 5-2. After the two paths of light are coupled into one beam through the 1*2 coupler 5-2, they enter the sensing optical fiber 6-2, and are reflected back to the 2*2 coupler 3-2 at the reflector 7-2 to form interference light, which is received by the photodetector 8-2. The output light of channel n of beam splitter 2 is divided into two paths of light through the 2*2 coupler 3-n, one of which enters the delay optical fiber 4-n and is connected to the 1*2 coupler 5-n; the other is connected to the other input port of the 1*2 coupler 5-n. After the two paths of light are coupled into one beam through the 1*2 coupler 5-n, they enter the sensing optical fiber 6-n, and are reflected back to the 2*2 coupler 3-n at the reflector 7-n to form interference light, which is received by the photodetector 8-n.

[0068] The length of the delay optical fiber 4-1 will be automatically adjusted according to the baseband signal to reduce the influence of interference noise caused by environmental vibration and improve the signal-to-noise ratio.

[0069] Step 2: The detector 8-1 transmits the received signal to the signal processing system. The signal processing system extracts the effective vibration signal and extracts eigenvalue 1 and eigenvalue 2 according to the zero-crossing energy and the proportion of the vibration signal; extracts eigenvalue 3 and eigenvalue 4 according to the frequency domain signal band variance and spectral entropy; calculates the Mel coefficient and extracts eigenvalues ​​5 to 8.

[0070] Step 3: Organize the eigenvalues ​​into eigenvalue vectors.

[0071] Step 4: Collect multiple groups of eigenvalue vectors of different types of vibration signals to form a vibration signal recognition model database.

[0072] As shown in Figure 5, the original vibration signals of three different types of intrusion behaviors are undisturbed, raining, and railing shaking. As shown in Figure 5(a), Figure 5(b), and Figure 5(c), different types of intrusion behaviors at different times have completely different time domain vibration signals.

[0073] Step 6: Bring the vibration signal recognition model database into the SVM for training to generate a vibration signal recognition model.

[0074] like Figure 6 The figure shows the result of signal classification of different types of intrusion behaviors using feature 3 and feature 4. As can be seen from the figure, although the time domain signals are obviously different, the vibration signals generated by these intrusions can be classified through the special zone feature values, so that different types of intrusion behaviors can be effectively identified.

[0075] Secondly, the vibration signal recognition model generated is used to identify the vibration signal collected in real time through the above steps to determine the intrusion behavior, including:

[0076] Step 1: The continuous light output by light source 1 is split into multiple parts by beam splitter 2:

[0077] Step 2: The output light of channel 1 of beam splitter 2 is divided into two paths through 2*2 coupler 3-1, one of which enters the delay optical fiber 4-1 and is connected to 1*2 coupler 5-1; the other is connected to the other input port of 1*2 coupler 5-1. The two paths of light are coupled into one beam through 1*2 coupler 5-1 and enter the sensing optical fiber 6-1, and are reflected back to 2*2 coupler 3-1 at reflector 7-1 to form interference light, which is received by photodetector 8-1. The output light of channel 2 of beam splitter 2 is divided into two paths of light through 2*2 coupler 3-2, one of which enters the delay optical fiber 4-2 and is connected to 1*2 coupler 5-2; the other is connected to the other input port of 1*2 coupler 5-2. After the two paths of light are coupled into one beam through the 1*2 coupler 5-2, they enter the sensing optical fiber 6-2, and are reflected back to the 2*2 coupler 3-2 at the reflector 7-2 to form interference light, which is received by the photodetector 8-2. The output light of channel n of beam splitter 2 is divided into two paths of light through the 2*2 coupler 3-n, one of which enters the delay optical fiber 4-n and is connected to the 1*2 coupler 5-n; the other is connected to the other input port of the 1*2 coupler 5-n. After the two paths of light are coupled into one beam through the 1*2 coupler 5-n, they enter the sensing optical fiber 6-n, and are reflected back to the 2*2 coupler 3-n at the reflector 7-n to form interference light, which is received by the photodetector 8-n.

[0078] Step 3: The detector transmits the received signal to the signal processing system. The signal processing system extracts the effective vibration signal and extracts eigenvalue 1 and eigenvalue 2 according to the zero-crossing energy and the proportion of the vibration signal; extracts eigenvalue 3 and eigenvalue 4 according to the frequency domain signal band variance and spectral entropy; calculates the Mel coefficient and extracts eigenvalues ​​5 to 8.

[0079] Step 4: Organize the eigenvalues ​​into eigenvalue vectors.

[0080] Step 5: Bring the eigenvalue vector into SVM, use the vibration signal recognition model to identify the signal, and classify it according to the similarity matching between the eigenvector of the real-time collected signal and the eigenvector of each type of vibration data in the vibration signal recognition model to determine the intrusion behavior.

[0081] In summary, the intelligent fiber optic perimeter alarm device with adjustable vibration sensitivity designed by the present invention can adjust the sensitivity of vibration signals on the hardware platform, has the ability to identify vibration signals with intelligent low hardware platform requirements, is safe and reliable, and has the advantages of 24-hour uninterrupted intelligent monitoring. The intelligent fiber optic perimeter alarm method with adjustable vibration sensitivity, after collecting vibration signals through the intelligent fiber optic perimeter alarm device with adjustable vibration sensitivity, extracts 8 eigenvalues ​​such as time domain, frequency domain and Mel coefficient from the collected signals to form a eigenvector, and uses the support vector machine SVM for discrimination. By extracting eigenvalues ​​from a variety of different types of vibration signals, a vibration signal pattern library is formed, which can effectively identify and distinguish between rain, shaking railings, and no disturbance.

[0082] The above embodiments are merely descriptions of preferred implementation modes of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. An intelligent optical fiber perimeter alarm device with adjustable vibration sensitivity, characterized in that: The invention comprises a signal processing system, a light source (1), a beam splitter (2), and a plurality of parallel optical links connected to the beam splitter (2) via a communication optical fiber (9); each group of optical links comprises a 2*2 coupler (3), a delay optical fiber (4), and a 1*2 coupler (5) connected in sequence via the communication optical fiber (9); each group of optical links also comprises a reflector (7) connected to the 1*2 coupler (5) via a sensing optical fiber (6), and a detector (8) connected to the 2*2 coupler (3); the output light of the light source (1) is divided into a plurality of beams via the beam splitter (2), and each beam of light passes through the communication optical fiber (9) and then , is divided into two paths of light through a 2*2 coupler (3), one of which enters the delay optical fiber (4) and is connected to the 1*2 coupler (5); the other path is directly connected to the other input port of the 1*2 coupler (5); the two paths of light are coupled into a beam of light through the 1*2 coupler (5) and then enter the sensing optical fiber (6), and are reflected back to the 2*2 coupler (3) at the reflector (7) to form interference light, which is received by a photoelectric detector (8). The detector (8) performs photoelectric conversion on the received interference light and transmits it to the signal processing system to extract signal feature values ​​and perform identification to determine whether to issue an alarm signal.

2. The intelligent optical fiber perimeter alarm device with adjustable vibration sensitivity according to claim 1, characterized in that: The wavelength of the light source (1) ranges from visible light to infrared light, preferably 1310 to 1650 nm.

3. The intelligent optical fiber perimeter alarm device with adjustable vibration sensitivity according to claim 1, characterized in that: The number of channels of the beam splitter (2) may be 1 to 256.

4. The intelligent optical fiber perimeter alarm device with adjustable vibration sensitivity according to claim 1, characterized in that: The length of the communication optical fiber (9) ranges from 0.1m to 100km, the length of the delay optical fiber (4) ranges from 1m to 20km, and the length of the sensing optical fiber (6) ranges from 0.1m to 200km.

5. The intelligent optical fiber perimeter alarm device with adjustable vibration sensitivity according to claim 1, characterized in that: The beam splitter (2), 2*2 coupler (3), and 1*2 coupler (5) use equal-ratio beam splitting or non-equal-ratio beam splitting.

6. The intelligent optical fiber perimeter alarm device with adjustable vibration sensitivity according to claim 1, characterized in that: The reflectivity of the reflector (7) ranges from 1% to 100%.

7. An intelligent optical fiber perimeter alarm method with adjustable vibration sensitivity, using the intelligent optical fiber perimeter alarm device with adjustable vibration sensitivity as claimed in any one of claims 1 to 6, characterized in that: The following steps are involved: S1: Establish a vibration signal recognition model, including: S101: The beam splitter (2) channel divides the output light of the light source (1) into two paths of light through the 2*2 coupler (3), one of which enters the delay optical fiber (4) and is connected to the 1*2 coupler (5); the other path is connected to the other input port of the 1*2 coupler (5); the two paths of light are coupled into a beam of light through the 1*2 coupler (5) and then enter the sensing optical fiber (6), and are reflected back to the 2*2 coupler (3) at the reflector (7) to form interference light, which is received by the photodetector (8); S102: The detector (8) transmits the received signal to the signal processing system, and the signal processing system extracts the effective vibration signal, and extracts eigenvalue 1 and eigenvalue 2 according to the zero-crossing energy and the vibration signal ratio; extracts eigenvalue 3 and eigenvalue 4 according to the frequency domain signal band variance and spectral entropy; calculates the Mel coefficient, and extracts eigenvalues ​​5 to 8; S103: Combining the above eigenvalues ​​into an eigenvalue vector; S104: collecting multiple groups of eigenvalue vectors of different types of vibration signals to form a vibration signal recognition model database; S105: Bringing the vibration signal recognition model database into the SVM for training to generate a vibration signal recognition model; S2: Using the generated vibration signal recognition model to recognize the vibration signal collected in real time through the above steps, and judging the intrusion behavior, including: S201: adopting the above steps S101-S103, collecting the eigenvalues ​​of the vibration signal in real time, and forming a real-time signal eigenvalue vector; S202: Bring the above-mentioned real-time signal eigenvalue vector into SVM, use the vibration signal recognition model to identify the signal, and classify it according to the similarity matching between the eigenvector of the real-time collected signal and the eigenvectors of each type of vibration data in the vibration signal recognition model to determine the intrusion behavior.

Images