A perimeter, enclosure and physical protection intrusion linear alarm detection device

Abstract

The application provides a perimeter, enclosure and physical protection intrusion linear alarm identification device, characterized by comprising a monitoring center layer, a field detection layer and a network transmission layer, the field detection layer is responsible for collecting various types of intrusion related detection signals in the perimeter or the enclosure range, the detection signals are transmitted through the network transmission layer and then transmitted to the monitoring center layer, the monitoring center layer centrally monitors, analyzes and identifies the received detection signals to determine whether there is a real intrusion behavior; the monitoring center layer can synchronously trigger a subsequent alarm and disposal process, combined with the linkage function of a video monitoring subsystem, provides intuitive video evidence for the intrusion behavior, realizes rapid response and on-site driving away of the intrusion event, provides strong support for subsequent event tracing, and improves the closed-loop management capability of perimeter or enclosure security.

Description

Technical Field

[0001] This invention relates to the field of security system technology, and in particular to a perimeter, enclosure, and physical protection intrusion linear alarm and identification device. Background Technology

[0002] Currently, traditional perimeter or enclosure / physical protection security systems mostly employ technologies such as active infrared detection, microwave sensing, vibration sensing, leaky cable detection, electrostatic induction, and video intelligent recognition. However, under complex and variable weather conditions and environmental interference such as strong winds, heavy rain, hail, snowfall, and sandstorms, each single sensing method has significant limitations. These limitations are generally manifested in high false alarm and false alarm rates, making it difficult to guarantee system stability and reliability. Consequently, they cannot meet the stringent requirements for perimeter or enclosure / physical protection in high-security scenarios such as prisons, nuclear power plants, important warehouses, and military restricted areas.

[0003] Existing improvement solutions still largely rely on single sensing technologies, such as using only fiber optic vibration sensors or traditional vibration cables for intrusion detection. While these solutions improve environmental adaptability to some extent, they still struggle to effectively distinguish between genuine intrusion behavior and environmental interference signals in complex scenarios such as strong winds, swaying vegetation, small animal intrusions, and ground vibrations. False alarms and missed alarms remain unresolved, and overall security effectiveness still has significant shortcomings. Summary of the Invention

[0004] To address the aforementioned technical issues, this invention provides a perimeter, enclosure, and physical protection intrusion linear alarm identification device. This device integrates distributed fiber optic sensing technology and MEMS vibration sensing technology, combined with intelligent algorithms and video surveillance. It is capable of operating 24 / 7, intelligently distinguishing between real intrusions and environmental interference such as wind, rain, and small animals, and possesses high environmental adaptability and a low false alarm rate.

[0005] This invention provides a perimeter, enclosure, and physical protection intrusion linear alarm identification device, characterized by comprising a monitoring center layer, a field detection layer, and a network transmission layer. The field detection layer is responsible for collecting various intrusion-related detection signals within the perimeter, enclosure, or physical protection area. After being transmitted through the network transmission layer, the detection signals are sent to the monitoring center layer, where the monitoring center layer centrally monitors, analyzes, and identifies the received detection signals to determine whether there is a real intrusion behavior, and simultaneously triggers subsequent alarm and handling procedures.

[0006] The field detection layer includes a distributed vibration fiber optic detection subsystem, a MEMS detection subsystem, and an auxiliary alarm subsystem; The distributed vibration fiber optic detection subsystem includes a sensing cable that is laid along the perimeter or enclosure. One end of the sensing cable is connected to the fiber optic alarm host. The distributed vibration fiber optic detection subsystem is used to detect environmental vibration signals in real time and generate a raw sensing signal. The system server identifies the signal and then integrates it with environmental information to generate an alarm signal. The MEMS detection subsystem includes a 6-axis MEMS sensor, which is evenly distributed on the perimeter or fence to detect the vibration and attitude changes of the fence in real time. The MEMS detection subsystem is combined with a neural network algorithm to filter out environmental interference factors. The auxiliary alarm subsystem includes auxiliary lighting equipment, broadcasting equipment, and audible and visual alarm equipment. The auxiliary lighting equipment is installed on monitoring nodes along the perimeter or perimeter, or along the fence. The broadcasting equipment is installed on monitoring nodes along the perimeter or perimeter, and the audible and visual alarm equipment is installed on monitoring nodes along the perimeter or perimeter. Each device receives control signals from the alarm server to issue on-site warnings and audible and visual alarms to drive away detected intrusions.

[0007] The distributed vibration fiber optic detection subsystem adopts phase-sensitive optical time-domain reflectometer distributed fiber optic sensing technology.

[0008] The monitoring center layer includes an intelligent algorithm subsystem and a video surveillance subsystem. The intelligent algorithm subsystem is built into the system server. The intelligent algorithm subsystem includes an empirical mode decomposition module, a neural network recognition module, and a convolutional neural network module. It is used to perform in-depth analysis and feature extraction on the vibration signals transmitted by the distributed vibration fiber optic detection subsystem and the MEMS detection subsystem, intelligently distinguish between intrusion behavior and environmental interference, and ensure the accuracy of alarms. The specific steps for analysis and feature extraction are as follows: The first step is for the empirical mode decomposition module to preprocess the received raw detection signal, decompose the complex vibration signal mixed with environmental interference, achieve signal denoising and preliminary feature separation, remove the modal components corresponding to high-frequency random interference, and retain the effective signal components related to intrusion behavior. The second step is to input the effective signal components preprocessed by the empirical mode decomposition module into the neural network recognition module. This module learns autonomously from a large number of intrusion behavior and environmental interference samples, extracts the time domain and frequency domain features of the signal, performs preliminary classification of the effective signal, filters out suspected intrusion signals, and eliminates obvious environmental interference signals. The third step involves the convolutional neural network module extracting deep features and accurately identifying suspected intrusion signals after initial screening. It captures deep features of the signals through convolutional and pooling layers, and combines them with a pre-set intrusion behavior feature library to further distinguish suspected signals and accurately differentiate between real intrusion behavior and remaining minor environmental interference. The video surveillance subsystem uses high-definition cameras and thermal imaging equipment deployed along the perimeter or fence to monitor the perimeter or fence environment in real time. The system is linked with an intelligent alarm algorithm, and after an alarm is triggered, it automatically calls up video resources to provide intuitive video evidence of intrusion.

[0009] The network transmission layer includes a network transmission and supporting subsystems, which are used to transmit all data to the monitoring center layer to realize command and information interaction. The network transmission and supporting subsystems adopt fiber optic ring network technology.

[0010] The specific steps of the MEMS detection subsystem in filtering environmental interference factors using neural network algorithms include: Step 1: Signal Acquisition and Preprocessing: The 6-axis MEMS sensor synchronously acquires the vibration acceleration and angular velocity data of the fence at a sampling frequency of no less than 1kHz. After bandpass filtering and adaptive variational mode decomposition (VMD), the signal is decomposed into two types: information-dominated IMF and noise-dominated IMF. Step 2: Differentiated denoising: Short-window time-frequency peak filtering (TFPF) is used for information-dominant IMF to retain useful signal details, while long-window TFPF is used for noise-dominant IMF to strongly suppress environmental noise. The reconstructed signal is then denoised. Step 3: Feature extraction and neural network inference: Extract time-domain statistical features, frequency-domain energy features and time-frequency joint features from the denoised signal, construct multi-dimensional feature vectors and input them into a pre-trained CNN-LSTM hybrid neural network to identify vibration behavior patterns (normal / wind / rain / climbing / shearing / lifting, etc.). Step 4: Dynamic Threshold Decision and Adaptive Filtering: Based on the confidence level of the behavior category output by the neural network, statistical smoothing is performed using the Dynamic Mean Algorithm (DMA). The final alarm decision is made through multi-level adaptive thresholds, and false alarm samples are fed back to the neural network for online learning and optimization to continuously improve the ability to filter environmental interference.

[0011] The beneficial effects of this invention are as follows: 1. The monitoring center layer, field detection layer, and network transmission layer are linked together. The field detection layer can comprehensively collect dual-source detection signals from optical fiber and MEMS. The network transmission layer ensures real-time and uninterrupted signal transmission. The monitoring center layer completes centralized identification and alarm triggering through intelligent algorithms. Each layer has a clear division of labor and works together efficiently to ensure the stability and real-time responsiveness of the overall system operation and avoid system failure due to failure of a single link. 2. By linking three modules—Empirical Mode Decomposition (EMD), Neural Network Recognition, and Convolutional Neural Network (CNN)—the original detection signal is first denoised and purified, and then the signal is gradually classified and deeply discriminated. This can not only eliminate common environmental interferences such as strong winds, rain, and vegetation swaying, but also effectively distinguish between real intrusion behavior and minor interferences such as small animal disturbances and slight ground vibrations, greatly reducing the false alarm rate and false negative rate, and meeting the reliability requirements of high-security scenarios. 3. After the intelligent algorithm module completes intrusion identification, the monitoring center can simultaneously trigger subsequent alarms and handling processes. Combined with the linkage function of the video surveillance subsystem, it provides intuitive video evidence of intrusion behavior, which not only realizes rapid response and on-site expulsion of intrusion events, but also provides strong support for subsequent event tracing and improves the closed-loop management capability of perimeter or perimeter security. Attached Figure Description

[0012] Figure 1 This is a flowchart of the overall modules of the present invention; Figure 2 This is a schematic diagram of the fence installation of the 6-axis MEMS sensor of the present invention. Detailed Implementation

[0013] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0014] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for 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. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0015] 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 can understand the specific meaning of the above terms in this invention based on the specific circumstances. Furthermore, the technical features involved in the different embodiments of this invention described below can be combined with each other as long as they do not conflict with each other. Example 1

[0016] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings.

[0017] In this embodiment, a perimeter, enclosure and physical protection intrusion linear alarm identification device is deployed in a layered architecture. The field detection layer continuously lays Φ-OTDR distributed sensing optical cables along the perimeter or enclosure fence (adopting a combination of buried and attached fence deployment, with positioning accuracy ≤5 meters). At the same time, high-precision 6-axis MEMS sensors are installed at key nodes of the fence (one every 3 meters) to realize comprehensive acquisition of fence vibration, attitude change and surrounding ground vibration. The network transmission layer adopts a dual-fiber ring network redundancy design to transmit the signals collected by the field detection layer to the monitoring center layer in real time with a transmission delay of ≤300ms, ensuring uninterrupted signal transmission. The monitoring center layer deploys intelligent algorithm servers, monitoring terminals, and alarm terminals, and has built-in core algorithm modules such as Empirical Mode Decomposition (EMD), Neural Network Recognition, and Convolutional Neural Network (CNN), as well as an intrusion behavior feature library (covering signal feature samples of intrusion behaviors such as climbing, cutting, impacting, and digging).

[0018] The distributed vibration fiber optic detection subsystem of the on-site detection layer captures vibration signals in real time within the perimeter, enclosure, or physical protection area, including vibrations from people climbing fences, cutting fences, ground excavation, and environmental disturbance vibrations caused by strong winds, rain, and swaying vegetation. The MEMS detection subsystem synchronously detects changes in the fence's posture (such as deformation caused by shaking or pulling) and vibration signals. Both types of detection signals are transmitted in real time to the intelligent algorithm module in the monitoring center layer via the fiber optic ring network of the network transmission layer to complete the initial signal reception and buffering.

[0019] The EMD module preprocesses the received raw mixed vibration signal, decomposing the signal into 8 intrinsic mode functions (IMFs) and 1 residual component. The high-frequency IMF components, which correspond to random interference caused by strong winds and rainfall, are removed, while the effective mid- and low-frequency signal components related to intrusion behavior (such as continuous vibration signals generated by climbing fences and high-frequency pulse vibration signals generated by cutting fences) are retained. The preprocessed effective signal components are input into the neural network recognition module. This module calls the preset sample library to extract the time domain features (vibration duration and peak amplitude) and frequency domain features (frequency range) of the signal, performs preliminary classification of the signal, filters out suspected intrusion signals (such as vibration signals with a duration ≥3s and an amplitude ≥0.5g), and excludes obvious environmental interference signals (such as short-term, low-amplitude vibration signals caused by rainfall). The CNN module performs deep processing on suspected intrusion signals, extracting deep waveform features of the signals through convolutional layers and comparing them with an intrusion behavior feature database. For example, it compares suspected signals with the feature waveforms of climbing behavior. If the matching degree is ≥90%, it is determined to be a real intrusion behavior; if the matching degree is ≥85% with the feature waveforms of small animal disturbance, it is determined to be a minor interference and no alarm is triggered.

[0020] When the CNN module determines that a real intrusion has occurred, the monitoring center immediately triggers an alarm process, which is then linked to the video surveillance subsystem. The system automatically calls the high-definition cameras (supporting infrared night vision) and thermal imaging equipment in the alarm area to capture and record the intrusion scene in real time. At the same time, the auxiliary lighting, broadcasting and audible and visual alarm subsystems automatically turn on the LED auxiliary lighting, the audible and visual alarm equipment emits an alarm sound of ≥110dB, and the broadcasting equipment plays warning voice messages in a loop to drive people away from the scene. After monitoring center staff view the on-site footage through the monitoring terminal and confirm the intrusion situation, they can issue handling instructions to complete the closed-loop handling of the intrusion incident.

[0021] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. The various components mentioned in this invention are common technologies in the existing field. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A perimeter, enclosure, and physical protection intrusion linear alarm and identification device, characterized in that... It includes a monitoring center layer, a field detection layer, and a network transmission layer. The field detection layer is responsible for collecting various intrusion-related detection signals from the perimeter or within the perimeter area. After being transmitted through the network transmission layer, the detection signals are sent to the monitoring center layer, where the monitoring center layer centrally monitors, analyzes, and identifies the received detection signals to determine whether there is a real intrusion behavior, and simultaneously triggers subsequent alarms and handling procedures.

2. The perimeter, enclosure, and physical protection intrusion linear alarm identification device according to claim 1, characterized in that... The field detection layer includes a distributed vibration fiber optic detection subsystem, a MEMS detection subsystem, and an auxiliary alarm subsystem; The distributed vibration fiber optic detection subsystem includes a sensing cable that is laid along the perimeter or enclosure. One end of the sensing cable is connected to the fiber optic alarm host. The distributed vibration fiber optic detection subsystem is used to detect environmental vibration signals in real time and generate a raw sensing signal. The system server identifies the signal and then integrates it with environmental information to generate an alarm signal. The MEMS detection subsystem includes a 6-axis MEMS sensor, which is evenly distributed on the perimeter or fence to detect the vibration and attitude changes of the fence in real time. The MEMS detection subsystem is combined with a neural network algorithm to filter out environmental interference factors. The auxiliary alarm subsystem includes auxiliary lighting equipment, broadcasting equipment, and audible and visual alarm equipment. The auxiliary lighting equipment is installed on monitoring nodes along the perimeter or perimeter, or along the fence. The broadcasting equipment is installed on monitoring nodes along the perimeter or perimeter, and the audible and visual alarm equipment is installed on monitoring nodes along the perimeter or perimeter. Each device receives control signals from the alarm server to issue on-site warnings and audible and visual alarms to drive away detected intrusions.

3. The perimeter, enclosure, and physical protection intrusion linear alarm identification device according to claim 2, characterized in that... The distributed vibration fiber optic detection subsystem adopts phase-sensitive optical time-domain reflectometer distributed fiber optic sensing technology.

4. The perimeter, enclosure, and physical protection intrusion linear alarm identification device according to claim 2, characterized in that... The monitoring center layer includes an intelligent algorithm subsystem and a video surveillance subsystem. The intelligent algorithm subsystem is built into the system server. The intelligent algorithm subsystem includes an empirical mode decomposition module, a neural network recognition module, and a convolutional neural network module. It is used to perform in-depth analysis and feature extraction on the vibration signals transmitted by the distributed vibration fiber optic detection subsystem and the MEMS detection subsystem, intelligently distinguish between intrusion behavior and environmental interference, and ensure the accuracy of alarms. The specific steps for analysis and feature extraction are as follows: The first step is for the empirical mode decomposition module to preprocess the received raw detection signal, decompose the complex vibration signal mixed with environmental interference, achieve signal denoising and preliminary feature separation, remove the modal components corresponding to high-frequency random interference, and retain the effective signal components related to intrusion behavior. The second step is to input the effective signal components preprocessed by the empirical mode decomposition module into the neural network recognition module. This module learns autonomously from a large number of intrusion behavior and environmental interference samples, extracts the time domain and frequency domain features of the signal, performs preliminary classification of the effective signal, filters out suspected intrusion signals, and eliminates obvious environmental interference signals. The third step involves the convolutional neural network module extracting deep features and accurately identifying suspected intrusion signals after initial screening. It captures deep features of the signals through convolutional and pooling layers, and combines them with a pre-set intrusion behavior feature library to further distinguish suspected signals and accurately differentiate between real intrusion behavior and remaining minor environmental interference. The video surveillance subsystem uses high-definition cameras and thermal imaging equipment deployed along the front perimeter or fence to monitor the perimeter or fence environment in real time. The system is linked with intelligent alarm algorithm software, and after an alarm is triggered, it automatically calls up video resources to provide intuitive video evidence of intrusion.

5. The perimeter, enclosure, and physical protection intrusion linear alarm identification device according to claim 2, characterized in that... The network transmission layer includes a network transmission and supporting subsystems, which are used to transmit all data to the monitoring center layer to realize command and information interaction. The network transmission and supporting subsystems adopt fiber optic ring network technology.

6. The perimeter, enclosure, and physical protection intrusion linear alarm identification device according to claim 2, characterized in that... The specific steps of the MEMS detection subsystem in filtering environmental interference factors using neural network algorithm software include: Step 1: Signal Acquisition and Preprocessing: The 6-axis MEMS sensor synchronously acquires the vibration acceleration and angular velocity data of the fence at a sampling frequency of no less than 1kHz. After bandpass filtering and adaptive variational mode decomposition (VMD) software algorithm, the signal is decomposed into two categories: information-dominated IMF and noise-dominated IMF. Step 2: Differentiated denoising: Short window time-frequency peak filtering (TFPF) algorithm is used for information-dominant IMF to retain useful signal details, while long window TFPF is used for noise-dominant IMF to strongly suppress environmental noise. The reconstructed signal is then denoised. Step 3: Feature extraction and neural network algorithm reasoning: Extract time-domain statistical features, frequency-domain energy features and time-frequency joint features from the denoised signal, construct a multi-dimensional feature vector input to a pre-trained CNN-LSTM hybrid neural network algorithm, and identify vibration behavior patterns (normal / wind / rain / climbing / shearing / lifting, etc.). Step 4: Dynamic Threshold Decision and Adaptive Filtering: Based on the confidence level of the behavior category output by the neural network algorithm software, statistical smoothing is performed using the dynamic mean algorithm (DMA). The final alarm decision is completed through multi-level adaptive thresholds, and false alarm samples are fed back to the neural network for online learning and optimization to continuously improve the ability to filter environmental interference.

Images