A method and device for generating behavior features for intrusion detection system

Abstract

This invention provides a method and apparatus for generating behavioral features of intrusion alarm system detection, relating to the field of civil aviation safety monitoring technology. The method includes: acquiring triaxial acceleration and pressure signals of an airport fence; calculating the deformation of the airport fence under standard tension based on an inertial motion acquisition model to obtain an equivalent stiffness coefficient; calculating the synthetic amplitude of the triaxial acceleration signals based on the equivalent stiffness coefficient, and locating the spatial position of the intrusion event at the current time point using the synthetic amplitude result to obtain the intrusion trajectory; extracting time-frequency domain features from the pressure signal, and comparing and analyzing the signal time-frequency domain features with pressure feature thresholds to obtain intrusion intensity features; and calculating spatiotemporal correlation feature parameters for the intrusion trajectory and intrusion intensity features at each time point based on the time series to generate intrusion behavior features. This invention solves the problems of the reliability of detection results and the effectiveness of security decision-making.

Description

Technical Field

[0001] This invention relates to the field of civil aviation safety monitoring technology, and more specifically, to a method and apparatus for generating behavioral characteristics for intrusion alarm system detection. Background Technology

[0002] In existing civil aviation safety monitoring technologies, security personnel typically simulate intrusion behaviors by climbing, cutting fences, and prying to test the response capabilities of fence alarm systems. However, existing testing methods can only record a single qualitative result—whether the system triggers an alarm—without simultaneously collecting quantitative data such as the magnitude of the intrusion force, the actual deformation of the fence, changes in vibration frequency, and the specific movement path of the intrusion. When multiple simulated intrusion tests are conducted on the same fenced area, inconsistent test results often occur (sometimes an alarm, sometimes no alarm). Due to the lack of process records containing the aforementioned quantitative characteristics, inspectors cannot determine the root cause of this inconsistency: whether it is insufficient intrusion force, changes in the fence's physical state (such as stiffness), or a blind spot in the alarm system itself. This situation of "unable to distinguish the cause" leads to the loss of credibility of single test results, and subsequent security maintenance and protection strategy optimization become passive due to the lack of data, ultimately resulting in issues with the credibility of test results and the effectiveness of security decisions.

[0003] Therefore, there is an urgent need for a method and device for generating behavioral features for intrusion alarm system detection, which solves the problems of the reliability of detection results and the effectiveness of security decisions. Summary of the Invention

[0004] The purpose of this invention is to provide a method and apparatus for generating behavioral features for intrusion alarm system detection, thereby improving the aforementioned problems. To achieve the above objective, the technical solution adopted by this invention is as follows:

[0005] Firstly, this application provides a method for generating behavioral features for intrusion alarm system detection, including:

[0006] Acquire triaxial acceleration and pressure signals from the airport fence;

[0007] Based on the pre-set inertial motion acquisition model, the deformation of the field fence under standard tensile force is calculated to obtain the equivalent stiffness coefficient.

[0008] Based on the equivalent stiffness coefficient, the synthetic amplitude of the triaxial acceleration signal is calculated, and the spatial location of the intrusion event at the current time point is located by the synthetic amplitude result to obtain the intrusion trajectory.

[0009] The time-frequency domain features of the pressure signal are extracted, and the intrusion force features are obtained by comparing and analyzing the time-frequency domain features of the signal with a preset pressure feature threshold.

[0010] Based on a preset time series, spatiotemporal correlation feature parameters are calculated for the intrusion trajectory and intrusion intensity characteristics at each time point to generate intrusion behavior features.

[0011] Secondly, this application also provides a behavioral feature generation device for intrusion alarm system detection, comprising:

[0012] The acquisition module is used to acquire the triaxial acceleration and pressure signals of the airport fence;

[0013] The stiffness calculation module is used to calculate the deformation of the field fence under standard tensile force based on the preset inertial motion acquisition model, and obtain the equivalent stiffness coefficient.

[0014] The amplitude calculation module is used to perform synthetic amplitude calculation on the triaxial acceleration signal based on the equivalent stiffness coefficient, and to locate the spatial position of the intrusion event at the current time point through the synthetic amplitude result, thereby obtaining the intrusion trajectory.

[0015] The analysis module is used to extract the time-frequency domain features of the pressure signal, and compare and analyze the time-frequency domain features of the signal with a preset pressure feature threshold to obtain the intrusion force features;

[0016] The parameter calculation module is used to calculate the spatiotemporal correlation feature parameters of the intrusion movement trajectory and the intrusion intensity characteristics at each time point according to the preset time series, and generate intrusion behavior features.

[0017] The beneficial effects of this invention are as follows:

[0018] This invention overcomes the limitations of single-sensor detection by acquiring triaxial acceleration and pressure signals from airport fences in a dual-dimensional sensing manner. It comprehensively perceives fence intrusion behavior from both vibration and tactile force dimensions, making it more suitable for the large-scale monitoring needs of civil aviation airport perimeters. Based on this, an inertial motion acquisition model is used to calibrate the fence deformation under standard tension, and the equivalent stiffness coefficient is calculated. The equivalent stiffness coefficient is then used to correct and synthesize the amplitude, improving the spatial positioning accuracy of the intrusion point and the accuracy of motion trajectory reconstruction. Subsequently, the time-frequency domain characteristics of the signal are compared with pressure characteristic thresholds, overcoming the shortcomings of traditional simple time-domain discrimination and effectively distinguishing different intrusion intensity levels such as slight touch, deliberate prying, and forceful climbing. Finally, spatiotemporal correlation parameters are calculated based on the intrusion trajectory and intrusion intensity characteristics according to the time series, generating structured intrusion behavior features and achieving digital and standardized characterization of intrusion behavior. In summary, this invention solves the problems of the reliability of detection results and the effectiveness of security decisions.

[0019] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing embodiments of the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the behavioral feature generation method for intrusion alarm system detection described in this embodiment of the invention;

[0022] Figure 2 This is a schematic diagram of the behavioral feature generation device for the intrusion alarm system described in this embodiment of the invention.

[0023] The diagram is labeled as follows: 800, behavioral feature generation device for intrusion alarm system detection; 801, processor; 802, memory; 803, multimedia component; 804, I / O interface; 805, communication component. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0025] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0026] In the practical application environment of airport perimeter security inspection, when security personnel simulate intrusion behaviors to test the fence alarm system through methods such as climbing, cutting, and prying, the system often exhibits instability due to uneven fence stiffness distribution, environmental vibration interference, and the coupling effect of multiple factors such as intrusion force, speed, and location. This often results in the system alarming intermittently under the same tester and intrusion method. In such cases, traditional testing methods can only record qualitative results of whether an alarm is triggered, lacking the synchronous acquisition and recording of quantitative data such as fence deformation, vibration signal amplitude, pressure time-frequency characteristics, and intrusion trajectory. This makes it impossible for inspectors to distinguish whether the root cause of the inconsistent alarms is insufficient intrusion force, local fence stiffness changes, or a response blind spot in the alarm system itself. This renders single test results unreliable, and subsequent security decisions lack a basis.

[0027] Example 1:

[0028] This embodiment provides a method for generating behavioral features for intrusion alarm system detection.

[0029] See Figure 1 The figure shows that the method includes steps S1 to S5, including:

[0030] S1: Acquire the triaxial acceleration and pressure signals of the airport fence;

[0031] In this step, the triaxial acceleration signal and the pressure signal are acquired by multiple sets of acceleration and pressure sensors arranged in an array on the upper, middle and lower layers of the airport steel fence. The multiple sets of sensors are coupled to the intersection of the fence wires through magnetic bases to ensure the rigidity of the fence. When intrusion behaviors such as climbing, knocking, and squeezing occur, the multiple sets of sensors acquire the triaxial acceleration and pressure signals generated by the external force on the fence in real time at a sampling frequency of 1kHz. At the same time, the Beidou / GPS dual-mode receiver marks each frame of signal with a millisecond-level timestamp to ensure synchronous acquisition of multi-channel signals.

[0032] S2: Based on the pre-set inertial motion acquisition model, calculate the deformation of the field fence under standard tension to obtain the equivalent stiffness coefficient;

[0033] In this step, the inertial motion acquisition model includes an adjustable support, a deformation recording component (a rope-type displacement sensor), a force characteristic measurement component (an accelerometer connected to the signal processing module), a signal processing module, a storage module, a timing module, and a digital display device. The inertial motion acquisition model adopts a measurement method that combines static data and dynamic vibration data. The deformation recording component is used to record the deformation data of the steel fence to obtain static parameters characterizing the fence stiffness. The force characteristic measurement component is used to collect the dynamic vibration signal of the fence when it is subjected to external force. The timing module is used to provide a unified time reference so that the dynamic vibration signal corresponds one-to-one with time. The signal processing module processes the dynamic vibration signal in parallel, filters the contact point data corresponding to the maximum signal at each time point, and generates behavioral path visualization data based on the time series. The storage module is used to store static parameters, dynamic vibration signals, contact point data, and behavioral path data. The digital display device is used to display the above data in real time.

[0034] To clarify the specific method for obtaining the equivalent stiffness coefficient, step S2 includes S21 to S24, specifically:

[0035] S21: Obtain the standard tensile force value;

[0036] In this step, a fixed standard tensile force value conforming to the airport fence inspection specifications is collected using a standard tensile gauge with a range of 500N.

[0037] S22: Apply the standard tensile force value to multiple preset test points on the fence according to the inertial motion acquisition model, and select the target test points;

[0038] In this step, based on the inertial motion acquisition model and the airport fence grid structure and support column distribution, preset test points are set at multiple representative locations such as the grid intersection in the middle of the fence. Standard tensile force values ​​are uniformly applied to each preset test point. Points that are affected by edge support interference or uneven force are eliminated. The remaining points are further screened to select points with stable deformation response and reliable data, and finally the target test points are obtained.

[0039] S23: Based on the inertial motion acquisition model, the deformation of the target test point under standard tension is acquired in real time to obtain real-time deformation data;

[0040] In this step, the cable-type displacement sensor in the inertial motion acquisition model is used as a deformation recording component. The hook at the end of the sensor cable is connected to the target test point. Under the continuous action of standard tension, the elastic deformation displacement of the fence is collected through the inertial motion acquisition model to obtain real-time deformation data.

[0041] S24: Based on Hooke's Law, the equivalent stiffness is calculated by combining the preset standard tensile force value and the real-time deformation data to obtain the equivalent stiffness coefficient.

[0042] In this step, the standard tensile force value, the original span of the fence, the equivalent stress area, and the real-time deformation data are calculated according to Hooke's Law to obtain the equivalent stiffness coefficient, which is used to characterize the inherent stiffness characteristics of this section of the fence.

[0043] The expression for the equivalent stiffness coefficient is:

[0044] ;

[0045] In the above formula, This is the equivalent stiffness coefficient. This is the standard tensile force value. The original span of the fence. For the equivalent stress area, This is real-time deformation data.

[0046] S3: Based on the equivalent stiffness coefficient, the synthetic amplitude of the triaxial acceleration signal is calculated, and the spatial location of the intrusion event at the current time point is located by the synthetic amplitude result to obtain the intrusion trajectory;

[0047] To clarify the specific method for obtaining the intrusion trajectory, step S3 includes S31 to S34, specifically:

[0048] S31: During an intrusion event, the synthesized amplitude result is compared with a preset synthesized amplitude threshold. When the time of the synthesized amplitude result is continuously exceeding the threshold, it is determined that there is an intrusion behavior at the current time point.

[0049] During the intrusion event, the triaxial acceleration signal is subjected to high-pass and low-pass filtering to remove environmental vibration and temperature drift noise. The vector magnitude is calculated from the filtered triaxial acceleration signal to obtain a synthesized amplitude result. This synthesized amplitude result is compared with a synthesized amplitude threshold, and a continuous time window (e.g., duration ≥ 1) is set. (milliseconds) When the synthesized amplitude result consistently exceeds the synthesized amplitude threshold within a continuous time window, a real intrusion is determined to have occurred within the current time window; if the duration for which the synthesized amplitude result exceeds the synthesized amplitude threshold is less than the set window length, it is considered an instantaneous interference or false triggering event, and no alarm is triggered. By judging through a continuous period exceeding the threshold, rather than a single point in time, the processing logic is ensured to be consistent with the actual physical process.

[0050] The composite amplitude threshold is determined jointly by the equivalent stiffness coefficient of the fence and the background noise at the site.

[0051] S32: Based on the intrusion behavior, construct an intrusion direction vector by combining the triaxial acceleration signal with the equivalent stiffness coefficient, and then find the intersection of the rays to obtain the coordinates of the intrusion intersection point;

[0052] To clarify the specific method for obtaining the coordinates of the intrusion intersection, step S32 includes S321 to S325, specifically: S321: Obtain the mesh stiffness characteristic parameters;

[0053] In this step, the mesh stiffness characteristic parameters are used to perform weighted correction on the proportional relationship, so that the same physical intrusion behavior corresponds to different axial proportional distributions on meshes with different stiffness.

[0054] S322: Based on the intrusion behavior, multiple axial ratio values ​​are obtained by calculating the proportional relationship between the axial component of the triaxial acceleration signal, the equivalent stiffness coefficient, and the mesh stiffness characteristic parameters;

[0055] In this step, the X, Y, and Z axial components of the triaxial acceleration signal are extracted based on the intrusion behavior. The equivalent stiffness coefficient is then used to normalize and proportionally convert each axial component to eliminate the interference of stiffness differences at different locations of the fence on the signal direction determination, resulting in multiple axial proportional values ​​that reflect the actual direction of external force.

[0056] S323: Construct an invasion direction vector based on the multiple axial ratio values ​​as components of a three-dimensional spatial vector;

[0057] In this step, the multiple axial ratio values ​​are respectively used as the corresponding components of the three-dimensional spatial vector. Based on the actual installation position of the sensor, an invasion direction vector is constructed to characterize the direction of the intrusion force and the trend of its action.

[0058] S324: Construct an intrusion behavior motion ray along the vector direction based on the intrusion direction vector and the detection position at the current time point;

[0059] In this step, starting from the actual detection position of the sensor at the current time point, an intrusion behavior motion ray is constructed by extending along the intrusion direction vector. The intrusion behavior motion ray is used to reflect the movement trend of the intrusion behavior.

[0060] S325: Solve for the intersection point of the intrusion behavior ray and the boundary of the detection area to obtain the coordinates of the intrusion intersection point.

[0061] In this step, the intersection points of the intrusion behavior motion ray and the detection area boundary are solved using a geometric analytical algorithm to obtain the spatial intersection point coordinates of the intrusion behavior at the current moment, which are then used as the intrusion intersection point coordinates.

[0062] S33: Verify the validity of the intrusion intersection coordinates according to the range of the preset detection area boundary. When the intersection coordinates are within the range, they are considered valid intrusion coordinates.

[0063] In this step, the coordinates of the intrusion intersection are compared and verified with the actual physical boundary of the enclosure and the detection area range according to the range of the preset detection area boundary. On the one hand, it is compared whether the coordinates of the intrusion intersection are within the range of the actual physical boundary of the enclosure, and on the other hand, it is compared whether the coordinates of the intrusion intersection are within the range of the preset detection area. At the same time, abnormal coordinates caused by algorithm calculation errors, environmental interference and other factors are eliminated, and only the coordinates that match the constraints of the actual physical boundary of the enclosure and the preset detection area range are retained as valid intrusion coordinates.

[0064] S34: Smooth correction is performed based on the intensity of the effective intrusion coordinates and the synthetic amplitude result of the adjacent ones, and the intrusion trajectory is fitted by a spatial curve.

[0065] In this step, the effective intrusion coordinates adjacent to each other are weighted and smoothed using the time series as the axis to reduce the fluctuation error of discrete coordinates. The trajectory curvature and direction are corrected by combining the intensity of the synthesized amplitude result. A spatial curve fitting algorithm is used to fit all discrete effective intrusion coordinate points into a continuous and smooth curve, which represents the motion law of the actual climbing, squeezing, and crossing intrusion actions.

[0066] S4: Extract the time-frequency domain features of the pressure signal, and compare and analyze the time-frequency domain features of the signal with a preset pressure feature threshold to obtain the intrusion force features;

[0067] To clarify the specific method for obtaining intrusion strength characteristics, step S4 includes S41 to S46, specifically:

[0068] S41: Acquire pressure signal samples;

[0069] In this step, the pressure signal samples are acquired in real time by pressure sensors that are rigidly coupled to the historical airport perimeter.

[0070] S42: Extract time-domain and frequency-domain features from the pressure signal to obtain a set of time-frequency domain features of the signal;

[0071] In this step, time-domain features such as peak value, mean, variance, duration, and rise time are extracted from the pressure signal, and frequency-domain features such as main frequency, spectral energy distribution, and harmonic components are extracted through fast Fourier transform to form a set of time-frequency domain features of the signal.

[0072] S43: Extract time-frequency domain features from the pressure signal samples of different intrusion types and different intrusion intensities as training feature samples;

[0073] In this step, pressure signal samples are collected and time-frequency domain features are extracted according to the intrusion types such as climbing, knocking, squeezing, cutting nets, and prying, as well as the intrusion intensity of light, medium and heavy attacks, to construct standardized training feature samples.

[0074] S44: Cluster the training feature samples under the same intrusion type according to the K-means clustering algorithm to obtain a comprehensive feature threshold set;

[0075] In this step, the training feature samples of different intrusion types and speed patterns under the same intrusion type are unsupervised clustered according to the K-means clustering algorithm. The intrusion behavior level clusters (light touch, pressing, rapid impact, slow pressure) are divided by feature distance, and the central feature value of each cluster is calculated to form a comprehensive feature threshold set of pressure-acceleration that includes different intrusion types, intrusion intensity levels, action speed levels and related spatial distribution characteristics.

[0076] The features used in clustering include not only the amplitude, rise time, and duration of the pressure signal, but also information such as the rate of change in the acceleration or pressure signal used for spatial positioning and velocity analysis.

[0077] S45: Calculate the distance between the signal time-frequency domain feature set and the comprehensive feature threshold set, and use the intrusion type and intensity level with the minimum distance as the current intrusion type and current intensity level;

[0078] In this step, the Euclidean distance is calculated between the feature centers in the signal time-frequency domain feature set and the comprehensive feature threshold set. By comparing all distance values, the intrusion type and intensity level with the smallest distance are determined as the intrusion type and intensity level at the current detection time.

[0079] S46: Based on the preset intensity quantization mapping relationship, convert the current intensity level and the current intrusion type into intrusion intensity quantization values, and output the intrusion intensity feature.

[0080] In this step, the force level and physical force value in the force quantification mapping relationship convert the current force level into a quantifiable continuous force value to obtain the current force level. This is then combined with the force magnitude, action mode, and intensity level in the current intrusion type to construct and output the intrusion force feature.

[0081] S5: Calculate the spatiotemporal correlation feature parameters of the intrusion trajectory and the intrusion intensity characteristics at each time point according to the preset time series, and generate intrusion behavior features.

[0082] To clarify the specific methods for obtaining intrusion behavior characteristics, step S5 includes S51 to S54, specifically:

[0083] S51: Based on the time series, perform spatiotemporal alignment and feature fusion on the intrusion trajectory and intrusion force features at each time point to obtain multidimensional spatiotemporal correlation parameters;

[0084] In this step, based on the time series, which is a unified high-precision timestamp of the BeiDou / GPS timing module, the intrusion intensity features (coordinates, position, displacement change, direction, trajectory slope) at the same time point are spatiotemporally aligned with the intrusion intensity features (intensity level, quantized intensity value, signal amplitude). At the same time, multi-dimensional information such as position, time, intensity, direction, and amplitude are fused to form multi-dimensional spatiotemporal correlation parameters. The multi-dimensional spatiotemporal correlation parameters include spatial coordinates, timestamp, intensity value, signal amplitude, and trajectory slope.

[0085] S52: Based on a preset threshold filtering rule, the multidimensional spatiotemporal correlation parameters are filtered to obtain optimized correlation feature parameters;

[0086] In this step, based on the threshold screening rules, which include the noise threshold, effective force lower limit, and reasonable coordinate range in the airport perimeter detection scenario, the multi-dimensional spatiotemporal correlation parameters are screened to remove abnormal parameters and invalid data caused by environmental interference, signal drift, and calculation errors, and retain the effective parameters that can truly reflect the intrusion behavior as optimized correlation feature parameters.

[0087] S53: The optimized correlation feature parameters are serialized and recombined according to the order of the time series to form a standard intrusion feature vector;

[0088] In this step, the optimized correlation feature parameters are arranged according to the order of the time series, the data dimensions and data formats of all parameters are unified, and a standard intrusion feature vector is constructed. The standard intrusion feature vector is a vector with fixed length, standardized structure, and can be directly used for analysis and identification.

[0089] S54: Verify the validity of the standard intrusion feature vector. When the verified standard intrusion feature vector meets the preset dimension requirements, output the intrusion behavior feature.

[0090] In this step, the dimensional integrity, data continuity, and spatiotemporal rationality of the standard intrusion feature vector are validated. It is ensured that the vector has no missing data, no logical errors, no out-of-bounds parameters, and fully meets the preset dimensional specifications. After successful validation, the standard intrusion feature vector is output as the final intrusion behavior feature.

[0091] Example 2:

[0092] This embodiment provides a behavioral feature generation device for intrusion alarm system detection, the device comprising:

[0093] The acquisition module is used to acquire the triaxial acceleration and pressure signals of the airport fence;

[0094] The stiffness calculation module is used to calculate the deformation of the field fence under standard tensile force based on the preset inertial motion acquisition model, and obtain the equivalent stiffness coefficient.

[0095] To clarify the specific methods for obtaining the stiffness calculation module, the following are included:

[0096] The acquisition unit is used to acquire standard tensile force values;

[0097] An application unit is used to apply the standard tensile force value to multiple preset test points on the fence according to the inertial motion acquisition model, and to filter out the target test points.

[0098] The acquisition unit is used to acquire the deformation of the target test point under standard tension in real time according to the inertial motion acquisition model, and obtain real-time deformation data.

[0099] The stiffness calculation unit is used to perform equivalent stiffness calculation based on Hooke's law, combined with the preset standard tensile force value and the real-time deformation data, to obtain the equivalent stiffness coefficient.

[0100] The amplitude calculation module is used to perform synthetic amplitude calculation on the triaxial acceleration signal based on the equivalent stiffness coefficient, and to locate the spatial position of the intrusion event at the current time point through the synthetic amplitude result, thereby obtaining the intrusion trajectory.

[0101] To clarify the specific method for obtaining the amplitude calculation module, the following are details:

[0102] The threshold comparison unit is used to compare the synthesized amplitude result with a preset synthesized amplitude threshold during an intrusion event. When the time of the synthesized amplitude result is continuously exceeding the threshold, it is determined that there is an intrusion behavior at the current time point.

[0103] The coordinate construction unit is used to determine the intrusion behavior, construct an intrusion direction vector by the triaxial acceleration signal and the equivalent stiffness coefficient, and then find the intersection of the rays to obtain the coordinates of the intrusion intersection point.

[0104] To clarify the specific methods for obtaining coordinate construction units, the following are included:

[0105] Obtain sub-units to acquire mesh stiffness characteristic parameters;

[0106] The calculation subunit is used to determine the intrusion behavior and calculate multiple axial ratio values ​​based on the proportional relationship between the axial component of the triaxial acceleration signal, the equivalent stiffness coefficient, and the mesh stiffness characteristic parameters.

[0107] The first construction subunit is used to construct an invasion direction vector based on multiple axial ratio values ​​as components of a three-dimensional spatial vector.

[0108] The second construction subunit is used to construct an intrusion behavior motion ray along the vector direction based on the intrusion direction vector and the detection position at the current time point;

[0109] The intersection point solving subunit is used to solve for the intersection points of the intrusion behavior motion ray and the boundary of the detection area to obtain the coordinates of the intrusion intersection point.

[0110] The verification unit is used to verify the validity of the intrusion intersection coordinates according to the range of the preset detection area boundary. When the intersection coordinates are within the range, they are considered valid intrusion coordinates.

[0111] The correction unit is used to perform smooth correction based on the intensity of the adjacent effective intrusion coordinates and the synthetic amplitude result, and to fit the intrusion trajectory through a spatial curve.

[0112] The analysis module is used to extract the time-frequency domain features of the pressure signal, and compare and analyze the time-frequency domain features of the signal with a preset pressure feature threshold to obtain the intrusion force features;

[0113] To clarify the specific methods for obtaining the analysis module, the following are included:

[0114] The pressure acquisition unit is used to acquire pressure signal samples;

[0115] The feature extraction unit is used to extract time-domain features and frequency-domain features from the pressure signal to obtain a set of time-frequency domain features of the signal.

[0116] The sample extraction unit is used to extract time-frequency domain features as training feature samples based on the pressure signal samples of different intrusion types and different intrusion intensities.

[0117] Clustering unit, used to cluster the training feature samples under the same intrusion type according to the K-means clustering algorithm to obtain a comprehensive feature threshold set;

[0118] The distance calculation unit is used to perform distance calculation on the signal time-frequency domain feature set and the comprehensive feature threshold set, and uses the intrusion type and intensity level with the minimum distance as the current intrusion type and current intensity level.

[0119] The conversion unit is used to convert the current intensity level and the current intrusion type into intrusion intensity quantization values ​​according to a preset intensity quantization mapping relationship, and output the intrusion intensity feature.

[0120] The parameter calculation module is used to calculate the spatiotemporal correlation feature parameters of the intrusion movement trajectory and the intrusion intensity characteristics at each time point according to the preset time series, and generate intrusion behavior features.

[0121] It should be noted that the specific manner in which each module performs its operation in the apparatus described in the above embodiments has been described in detail in the embodiments of the method, and will not be elaborated here.

[0122] Example 3:

[0123] Corresponding to the above method embodiments, this embodiment also provides a behavior feature generation device for intrusion alarm system detection. The behavior feature generation device for intrusion alarm system detection described below and the behavior feature generation method for intrusion alarm system detection described above can be referred to in correspondence with each other.

[0124] Figure 2 This is a block diagram illustrating a behavioral feature generation device 800 for intrusion alarm system detection, according to an exemplary embodiment. Figure 2 As shown, the behavior feature generation device 800 detected by the intrusion alarm system may include: a processor 801 and a memory 802. The behavior feature generation device 800 may also include one or more of a multimedia component 803, an I / O interface 804, and a communication component 805.

[0125] The processor 801 controls the overall operation of the behavior feature generation device 800 for intrusion alarm system detection to complete all or part of the steps in the behavior feature generation method for intrusion alarm system detection described above. The memory 802 stores various types of data to support the operation of the behavior feature generation device 800 for intrusion alarm system detection. This data may include, for example, instructions for any application or method operating on the behavior feature generation device 800 for intrusion alarm system detection, as well as application-related data such as contact data, sent and received messages, images, audio, video, etc. The memory 802 can be implemented using any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touchscreen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in the memory 802 or transmitted via the communication component 805. The audio component also includes at least one speaker for outputting audio signals. I / O interface 804 provides an interface between processor 801 and other interface modules, such as keyboards, mice, and buttons. These buttons can be virtual or physical. Communication component 805 is used for wired or wireless communication between the behavior feature generation device 800 detected by the intrusion alarm system and other devices. Wireless communication includes, for example, Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, or 4G, or a combination thereof. Therefore, the corresponding communication component 805 may include a Wi-Fi module, a Bluetooth module, or an NFC module.

[0126] In an exemplary embodiment, the behavior feature generation device 800 for intrusion alarm system detection may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the behavior feature generation method for intrusion alarm system detection described above.

[0127] Example 4:

[0128] Corresponding to the above method embodiments, this embodiment also provides a medium. The medium described below can be referred to in conjunction with the behavioral feature generation method for intrusion alarm system detection described above.

[0129] A medium storing a computer program, which, when executed by a processor, implements the steps of the behavioral feature generation method for intrusion alarm system detection described in the above method embodiments.

[0130] The medium can specifically be any medium capable of storing program code, such as a USB flash drive, external hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0131] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0132] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for generating behavioral features for intrusion alarm system detection, characterized in that, include: Acquire triaxial acceleration and pressure signals from the airport fence; Based on the pre-set inertial motion acquisition model, the deformation of the field fence under standard tensile force is calculated to obtain the equivalent stiffness coefficient. Based on the equivalent stiffness coefficient, the synthetic amplitude of the triaxial acceleration signal is calculated, and the spatial location of the intrusion event at the current time point is located by the synthetic amplitude result to obtain the intrusion trajectory. The time-frequency domain features of the pressure signal are extracted, and the intrusion force features are obtained by comparing and analyzing the time-frequency domain features of the signal with a preset pressure feature threshold. Based on a preset time series, spatiotemporal correlation feature parameters are calculated for the intrusion trajectory and intrusion intensity characteristics at each time point to generate intrusion behavior features.

2. The method for generating behavioral features for intrusion alarm system detection according to claim 1, characterized in that, Based on the pre-set inertial motion acquisition model, the deformation of the field fence under standard tensile force is calculated to obtain the equivalent stiffness coefficient, including: Obtain the standard tensile force value; According to the inertial motion acquisition model, the standard tensile force value is applied to multiple preset test points on the fence, and the target test points are selected by screening. The deformation of the target test point under standard tension is collected in real time according to the inertial motion acquisition model to obtain real-time deformation data. According to Hooke's Law, the equivalent stiffness is calculated by combining the preset standard tensile force value and the real-time deformation data to obtain the equivalent stiffness coefficient.

3. The method for generating behavioral features for intrusion alarm system detection according to claim 1, characterized in that, By synthesizing the amplitude results, the spatial location of the intrusion event at the current time point is determined, and the intrusion trajectory is obtained, including: During an intrusion event, the synthesized amplitude result is compared with a preset synthesized amplitude threshold. When the time of the synthesized amplitude result is continuously exceeding the threshold, it is determined that there is an intrusion behavior at the current time point. Based on the intrusion behavior, the intrusion direction vector is constructed by the triaxial acceleration signal and the equivalent stiffness coefficient, and the intersection of the ray is obtained to get the coordinates of the intrusion intersection point; The validity of the intrusion intersection coordinates is verified based on the range of the preset detection area boundary. When the intersection coordinates are within the range, they are considered valid intrusion coordinates. The intensity of the effective intrusion coordinates and the synthetic amplitude result are smoothed and corrected, and the intrusion trajectory is fitted by a spatial curve.

4. The method for generating behavioral features for intrusion alarm system detection according to claim 3, characterized in that, Based on the intrusion behavior, the intrusion direction vector is constructed by combining the triaxial acceleration signal with the equivalent stiffness coefficient, and the intersection of the vectors is calculated to obtain the coordinates of the intrusion intersection point, including: Obtain the stiffness characteristic parameters of the wire mesh; Based on the intrusion behavior, multiple axial ratio values ​​are obtained by calculating the proportional relationship between the axial component of the triaxial acceleration signal, the equivalent stiffness coefficient, and the mesh stiffness characteristic parameters. An invasion direction vector is constructed based on multiple axial ratio values ​​as components of a three-dimensional spatial vector. Based on the intrusion direction vector and the detection position at the current time point, an intrusion behavior motion ray is constructed along the vector direction; The intersection points of the intrusion behavior ray and the boundary of the detection area are calculated to obtain the coordinates of the intrusion intersection points.

5. The method for generating behavioral features for intrusion alarm system detection according to claim 1, characterized in that, The time-frequency domain features of the pressure signal are extracted, and the intrusion force features are obtained by comparing and analyzing the time-frequency domain features with a preset pressure feature threshold, including: Acquire pressure signal samples; The time-domain and frequency-domain features of the pressure signal are extracted respectively to obtain a set of time-frequency domain features of the signal; Time-frequency domain features are extracted from the pressure signal samples of different intrusion types and intrusion intensities as training feature samples; The training feature samples under the same intrusion type are clustered according to the K-means clustering algorithm to obtain a comprehensive feature threshold set; The distance between the signal time-frequency domain feature set and the comprehensive feature threshold set is calculated, and the intrusion type and intensity level with the minimum distance are used as the current intrusion type and current intensity level. Based on a preset intensity quantization mapping relationship, the current intensity level and the current intrusion type are converted into intrusion intensity quantization values, and the intrusion intensity characteristics are output.

6. A behavioral feature generation device for intrusion alarm system detection, characterized in that, include: The acquisition module is used to acquire the triaxial acceleration and pressure signals of the airport fence; The stiffness calculation module is used to calculate the deformation of the field fence under standard tensile force based on the preset inertial motion acquisition model, and obtain the equivalent stiffness coefficient. The amplitude calculation module is used to perform synthetic amplitude calculation on the triaxial acceleration signal based on the equivalent stiffness coefficient, and to locate the spatial position of the intrusion event at the current time point through the synthetic amplitude result, thereby obtaining the intrusion trajectory. The analysis module is used to extract the time-frequency domain features of the pressure signal, and compare and analyze the time-frequency domain features of the signal with a preset pressure feature threshold to obtain the intrusion force features; The parameter calculation module is used to calculate the spatiotemporal correlation feature parameters of the intrusion movement trajectory and the intrusion intensity characteristics at each time point according to the preset time series, and generate intrusion behavior features.

7. The behavior feature generation device for intrusion alarm system detection according to claim 6, characterized in that, The stiffness calculation module includes: The acquisition unit is used to acquire standard tensile force values; An application unit is used to apply the standard tensile force value to multiple preset test points on the fence according to the inertial motion acquisition model, and to filter out the target test points. The acquisition unit is used to acquire the deformation of the target test point under standard tension in real time according to the inertial motion acquisition model, and obtain real-time deformation data. The stiffness calculation unit is used to perform equivalent stiffness calculation based on Hooke's law, combined with the preset standard tensile force value and the real-time deformation data, to obtain the equivalent stiffness coefficient.

8. The behavior feature generation device for intrusion alarm system detection according to claim 6, characterized in that, The intrusion trajectory of the amplitude calculation module includes: The threshold comparison unit is used to compare the synthesized amplitude result with a preset synthesized amplitude threshold during an intrusion event. When the time of the synthesized amplitude result is continuously exceeding the threshold, it is determined that there is an intrusion behavior at the current time point. The coordinate construction unit is used to determine the intrusion behavior, construct an intrusion direction vector by the triaxial acceleration signal and the equivalent stiffness coefficient, and then find the intersection of the rays to obtain the coordinates of the intrusion intersection point. The verification unit is used to verify the validity of the intrusion intersection coordinates according to the range of the preset detection area boundary. When the intersection coordinates are within the range, they are considered valid intrusion coordinates. The correction unit is used to perform smooth correction based on the intensity of the adjacent effective intrusion coordinates and the synthetic amplitude result, and to fit the intrusion trajectory through a spatial curve.

9. The behavior feature generation device for intrusion alarm system detection according to claim 8, characterized in that, The coordinate construction unit includes: Obtain sub-units to acquire mesh stiffness characteristic parameters; The calculation subunit is used to determine the intrusion behavior and calculate multiple axial ratio values ​​based on the proportional relationship between the axial component of the triaxial acceleration signal, the equivalent stiffness coefficient, and the mesh stiffness characteristic parameters. The first construction subunit is used to construct an invasion direction vector based on multiple axial ratio values ​​as components of a three-dimensional spatial vector. The second construction subunit is used to construct an intrusion behavior motion ray along the vector direction based on the intrusion direction vector and the detection position at the current time point; The intersection point solving subunit is used to solve for the intersection points of the intrusion behavior motion ray and the boundary of the detection area to obtain the coordinates of the intrusion intersection point.

10. The behavior feature generation device for intrusion alarm system detection according to claim 6, characterized in that, The analysis module includes: The pressure acquisition unit is used to acquire pressure signal samples; The feature extraction unit is used to extract time-domain features and frequency-domain features from the pressure signal to obtain a set of time-frequency domain features of the signal. The sample extraction unit is used to extract time-frequency domain features as training feature samples based on the pressure signal samples of different intrusion types and different intrusion intensities. Clustering unit, used to cluster the training feature samples under the same intrusion type according to the K-means clustering algorithm to obtain a comprehensive feature threshold set; The distance calculation unit is used to perform distance calculation on the signal time-frequency domain feature set and the comprehensive feature threshold set, and uses the intrusion type and intensity level with the minimum distance as the current intrusion type and current intensity level. The conversion unit is used to convert the current intensity level and the current intrusion type into intrusion intensity quantization values ​​according to a preset intensity quantization mapping relationship, and output the intrusion intensity feature.

Images