Anti-uav comprehensive situation awareness system and method based on multi-source data fusion

Abstract

The application relates to the technical field of low-altitude defense, and discloses an anti-unmanned aerial vehicle comprehensive situation awareness system and method based on multi-source data fusion, which comprises a sensor acquisition unit, a data receiving and processing unit and a data analysis and situation generation unit. The sensor acquisition unit fuses radar, photoelectric and radio signals to obtain target original data; the data receiving and processing unit pre-processes the data integration; and the data analysis and situation generation unit performs identification and deduction based on a three-dimensional space situation. A monitoring object calibration and identification unit drives a comprehensive situation two-dimensional / three-dimensional display unit to realize visual output; and a data storage and management unit archives data to form a closed-loop management and control. Through the construction of a unified fusion platform, the integration of heterogeneous data and the combination of a geographic information model, high-precision, real-time, two-dimensional / three-dimensional integrated situation awareness and intelligent analysis of the low-altitude airspace around an airport are realized.

Description

Technical Field

[0001] This invention relates to the field of low-altitude defense technology, specifically to a comprehensive anti-drone situational awareness system and method based on multi-source data fusion. Background Technology

[0002] With the booming development of the low-altitude economy and the accelerated opening of urban airspace, the safety hazards posed by micro-drones are becoming increasingly prominent, especially in urban core areas with high-rise buildings and complex electromagnetic environments. Therefore, building an efficient integrated situational awareness system for countering drones has become an urgent need. Existing technologies typically employ a heterogeneous sensor network composed of low-altitude surveillance radar, radio spectrum detection, and photoelectric tracking equipment for collaborative detection, but many technical bottlenecks still exist in practical applications.

[0003] First, existing multi-source sensor data fusion technologies mostly employ static fusion strategies. Traditional multi-sensor fusion algorithms (such as standard Kalman filtering or information filtering) typically pre-determine the measurement noise covariance matrix and trust weights of each sensor during the system design phase. However, the detection performance of physical sensors is significantly affected by real-time weather conditions. For example, rain attenuation reduces the radar detection signal-to-noise ratio, and haze weakens the contrast of photoelectric imaging. Existing technologies lack a mechanism to dynamically adjust sensor weights based on real-time environmental meteorological parameters. This results in the inability to adaptively reduce the data weights of interfered sensors under severe or variable weather conditions, thus seriously affecting the accuracy and stability of target state estimation.

[0004] Secondly, existing counter-drone systems primarily rely on a geometric understanding of the monitoring environment, lacking a deep understanding of the physical properties of ground surfaces. In complex urban environments, the strong specular reflection of building glass curtain walls, electromagnetic scattering from metal surfaces, and thermal radiation from air conditioning units can easily create false signals resembling drone targets in radar echoes and infrared images. Current technologies struggle to effectively utilize semantic information about the geographical environment (such as material properties, reflectivity, and heat capacity) to help eliminate these static ground features with high radiation characteristics, resulting in a persistently high false alarm rate in complex urban settings.

[0005] Finally, the existing situation awareness systems exhibit a fragmented display and control model, lacking a closed-loop feedback mechanism. On one hand, traditional command terminals typically display two-dimensional electronic maps and three-dimensional video feeds separately, making it difficult for commanders to intuitively and synchronously correspond between macro-level situations and micro-level details, increasing cognitive load and command delays. On the other hand, existing systems usually operate in an open-loop manner, lacking effective human-machine interaction calibration and data archiving feedback processes. When false alarms occur, the system cannot easily receive confirmation or correction commands from operators to update the situation display in real time, nor can it utilize historical operational data for continuous system optimization, resulting in a failure to improve the system's long-term environmental adaptability. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a comprehensive anti-drone situational awareness system and method based on multi-source data fusion. It solves the problems of low fusion accuracy caused by inconsistent spatiotemporal benchmarks of multi-source heterogeneous data, the susceptibility of single sensors to environmental and meteorological factors leading to decreased target recognition rate, and the lack of an intuitive two-dimensional and three-dimensional integrated command view and manual closed-loop calibration mechanism.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] The first aspect of the present invention provides an anti-drone integrated situational awareness system, including a sensor acquisition unit, a data receiving and processing unit, a data analysis and situation generation unit, an integrated situational two-dimensional / three-dimensional display unit, a monitoring object calibration and identification unit, and a data storage and management unit.

[0009] The sensor acquisition unit is configured to acquire multi-dimensional physical observation data within the monitoring area, including a low-altitude surveillance radar, a radio spectrum detection device, and an optoelectronic tracking turntable; the low-altitude surveillance radar outputs a target point data sequence containing spatial position and motion parameters; the radio spectrum detection device outputs a detection data packet containing signal characteristics; and the optoelectronic tracking turntable outputs optoelectronic state vectors and video stream data in real time.

[0010] The data receiving and processing unit is connected to each device in the sensor acquisition unit and is configured to perform time synchronization and coordinate transformation processing. The unit sends a unified time pulse signal to each device, performs time alignment of multi-source data according to the timestamp, and converts heterogeneous data into a structured spatiotemporal dataset in the geocentric coordinate system.

[0011] The data storage and management unit stores three-dimensional geographic information data of the monitoring area, sensor configuration parameters, and full-link archived data; the three-dimensional geographic information data includes digital elevation models, oblique photogrammetry three-dimensional mesh models, and vectorized building outline data.

[0012] The data analysis and situation generation unit is configured to perform tasks such as loading the digital twin geographic base, multi-level data fusion, and identification and inference. This unit reads three-dimensional geographic information data to construct a digital twin geographic base under a unified coordinate system and physically defines the surface attributes of the three-dimensional mesh model. This unit dynamically adjusts the measurement noise covariance matrix of each sensor according to real-time meteorological environmental parameters and uses the optimal weighted fusion algorithm to calculate the comprehensive state vector of the target.

[0013] The integrated situational awareness 2D / 3D display unit is configured to perform integrated 2D / 3D situational awareness presentation; the unit includes a 2D electronic map rendering engine, a 3D digital twin rendering engine, and a 2D / 3D view linkage controller; the 2D / 3D view linkage controller calculates the spatial pose of the virtual camera in the 3D scene in real time based on the center point coordinates and scale level of the 2D view, realizing frame-level synchronization and field of view analysis of the 2D / 3D viewport.

[0014] The monitoring object calibration and identification unit is equipped with a human-machine interaction feedback interface, which is used to receive confirmation or correction instructions from operators for alarm targets, generate feedback data packets and send them to the data analysis and situation generation unit to update the situation display in real time, forming a closed-loop control.

[0015] The second aspect of the present invention provides a comprehensive situational awareness method for countering unmanned aerial vehicles (UAVs), including digital twin scene construction and physical attribute definition, spatiotemporal alignment and preprocessing of multi-source data, multi-level adaptive data fusion, two-dimensional and three-dimensional integrated situational awareness presentation, and calibration feedback and closed-loop control steps.

[0016] In the digital twin scene construction and physical attribute definition steps, two-dimensional vector geographic information data and oblique photogrammetry three-dimensional real scene data of the monitoring area are acquired and uniformly converted to the geocentric-ground-fixed coordinate system; semantic segmentation is performed on the real scene three-dimensional mesh model, and physical attribute feature vectors containing visible light diffuse reflectance, infrared thermal emissivity, specular reflectance coefficient and thermal capacity coefficient are generated for each triangular facet according to the material category, and geometric normal vectors are calculated to construct a digital twin geographic base with physical interactive computing capabilities.

[0017] In the spatiotemporal alignment and preprocessing steps of multi-source data, timestamp alignment is performed on radar, radio and optoelectronic data based on the system clock; a rotation transformation matrix is ​​constructed to convert the target point trace in the radar polar coordinate system into a three-dimensional position vector in the geocentric geofixed coordinate system; and a viewpoint transformation matrix based on the pinhole camera projection model is constructed to describe the mapping relationship between the three-dimensional spatial points of the optoelectronic target and the two-dimensional image pixel plane.

[0018] In the multi-level adaptive data fusion step, data layer clustering and association are performed on the spatiotemporally aligned multi-source data; real-time environmental parameters are obtained, and the interference sensitivity coefficients of environmental factors to each sensor are calculated, thereby dynamically adjusting the measurement noise covariance matrix of each sensor; an optimal weighted fusion algorithm based on an information filtering architecture is adopted, using the inverse matrix of the adaptive measurement noise covariance matrix as weighting coefficients to calculate the target comprehensive state vector containing position, velocity, and category.

[0019] In the integrated 2D / 3D situation presentation step, the 2D electronic map rendering engine is used to draw target tactical symbols and trajectories; the 3D digital twin rendering engine is used to render target models in a 3D scene with physical properties; the 2D view operation parameters are captured by the 2D / 3D view linkage controller, the virtual camera position vector and the line of sight are calculated, the 3D view matrix is ​​updated synchronously, and the line of sight occlusion is calculated in combination with 3D geometric data.

[0020] In the calibration feedback and closed-loop control steps, false alarm marking or type correction instructions for suspected targets are received through the human-machine interface, and feedback data packets are generated. False alarms are removed or target display is corrected in real time based on the feedback results, and the original monitoring data, fusion results and manual calibration labels are stored in the spatiotemporal index database to support the synchronous playback of historical situations and continuous system optimization.

[0021] This invention provides a comprehensive anti-drone situational awareness system and method based on multi-source data fusion. It has the following beneficial effects:

[0022] 1. This invention constructs a digital twin geographic base with physical interaction computing capabilities by performing semantic segmentation on oblique photogrammetry 3D real scene data and establishing a physical attribute feature vector containing visible light diffuse reflectance and infrared thermal emissivity parameters. The physical attribute data is used to assist multi-source sensors in distinguishing real targets from building background interference in complex urban backgrounds, thereby improving the identification capability for low, slow and small UAV targets.

[0023] 2. This invention acquires environmental meteorological parameters of the monitoring area in real time and calculates the interference sensitivity coefficients of environmental factors to each sensor. It uses an adaptive measurement noise covariance matrix to dynamically adjust the weight allocation of low-altitude surveillance radar, radio spectrum detection equipment and photoelectric tracking turntable in the multi-source fusion algorithm. Under variable weather conditions, it maintains the continuity and stability of target state estimation and solves the problem of detection failure caused by the performance degradation of a single sensor in harsh environments.

[0024] 3. This invention uses a two-dimensional and three-dimensional view linkage controller to calculate the virtual camera pose in real time to achieve viewport synchronization between the two-dimensional electronic map and the three-dimensional digital twin scene. Combined with the human-computer interaction interface, it receives calibration feedback instructions for false alarm targets and generates feedback data packets to update the situation display and archived data, providing command personnel with decision-making assistance that combines macro and micro perspectives and ensuring the accuracy of system operation data. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the overall architecture of the anti-drone integrated situational awareness system based on multi-source data fusion of the present invention;

[0026] Figure 2 This is a schematic diagram of the digital twin scene construction and geographic information preprocessing process of the present invention;

[0027] Figure 3 This is a schematic diagram of the multi-source data spatiotemporal alignment and preprocessing process of the present invention;

[0028] Figure 4 This is a schematic diagram of the multi-level data fusion and situation generation process of the present invention;

[0029] Figure 5 This is a schematic diagram of the integrated two-dimensional and three-dimensional situation presentation process of the present invention;

[0030] Figure 6 This is a schematic diagram of the calibration feedback and closed-loop control process of the present invention.

[0031] The system includes: 100, Sensor Acquisition Unit; 101, Low-Altitude Surveillance Radar; 102, Radio Spectrum Detection Equipment; 103, Optoelectronic Tracking Turntable; 1031, Precision Servo Gimbal; 1032, Visible Light Camera; 1033, Infrared Thermal Imager; 200, Data Receiving and Processing Unit; 300, Data Analysis and Situation Generation Unit; 400, Comprehensive Situation 2D / 3D Display Unit; 400a, 2D / 3D View Linkage Controller; 400b, 2D Electronic Map Rendering Engine; 400c, 3D Digital Twin Rendering Engine; 500, Monitoring Object Calibration and Identification Unit; and 600, Data Storage and Management Unit. Detailed Implementation

[0032] The technical solutions in 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 embodiments of the present invention, and not all embodiments. 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.

[0033] See attached document Figure 1 This invention provides a comprehensive anti-UAV situational awareness system, comprising a sensor acquisition unit 100, a data receiving and processing unit 200, a data analysis and situation generation unit 300, a comprehensive situational awareness two-dimensional / three-dimensional display unit 400, a monitoring object calibration and identification unit 500, and a data storage and management unit 600. The sensor acquisition unit 100 is configured to acquire multi-dimensional physical observation data within the monitoring area. The sensor acquisition unit 100 includes a low-altitude surveillance radar 101, a radio spectrum detection device 102, and an electro-optical tracking turntable 103.

[0034] The low-altitude surveillance radar 101 is mounted at an open, high point within the monitoring area. The low-altitude surveillance radar 101 is configured as either a fully coherent pulse-Doppler radar or a frequency-modulated continuous wave radar. The low-altitude surveillance radar 101 continuously scans the airspace within a 360-degree horizontal azimuth and vertical elevation range through mechanical rotation or electronic scanning. The low-altitude surveillance radar 101 is configured to output a sequence of target point data. Each point data point in the target point data sequence... It includes the target's spatial position parameters and motion parameters in the radar polar coordinate system.

[0035] Define the spot data output by the low-altitude surveillance radar 101 as follows:

[0036] ;

[0037] in: This indicates the straight-line slant range between the target and the phase center of the low-altitude surveillance radar 101. Indicates the azimuth angle of the target relative to the radar's north-facing reference; Indicates the elevation angle of the target relative to the radar horizontal reference; Indicates the radial velocity of the target along the radar line of sight; This indicates the signal-to-noise ratio of the echo signal.

[0038] The radio spectrum detection device 102 is deployed at the center or boundary of the monitoring area. The radio spectrum detection device 102 is connected to an omnidirectional antenna array or a directional antenna array. The radio spectrum detection device 102 operates in the frequency range of 300 MHz to 6000 MHz. The radio spectrum detection device 102 is configured to perform real-time scanning and analysis of the ambient electromagnetic spectrum. The radio spectrum detection device 102 outputs detection data packets containing signal characteristics. The detection data packets include the signal center frequency, signal bandwidth, signal modulation type, and received signal strength indication.

[0039] The photoelectric tracking turntable 103 includes a precision servo gimbal 1031, a visible light camera 1032, and an infrared thermal imager 1033. The visible light camera 1032 and the infrared thermal imager 1033 are rigidly fixed to the load platform of the precision servo gimbal 1031. The precision servo gimbal 1031 has dual-degree-of-freedom rotation capability along the azimuth and pitch axes. The precision servo gimbal 1031 is configured to rotate to a specified angle in response to external control commands. The visible light camera 1032 is configured to acquire video stream data in the visible light band. The infrared thermal imager 1033 is configured to acquire thermal radiation video stream data in the long-wave infrared band or the mid-wave infrared band. The photoelectric tracking turntable 103 outputs a photoelectric state vector in real time. .

[0040] Define the photoelectric state vector output by the photoelectric tracking turntable 103 as follows:

[0041] ;

[0042] in: This indicates the current azimuth angle of the precision servo gimbal 1031; This indicates the current pitch angle of the precision servo gimbal 1031; This indicates the current lens focal length value of the visible light camera 1032 or the infrared thermal imager 1033. This indicates the rotational angular velocity of the azimuth axis of the 1031 precision servo gimbal; This indicates the rotational angular velocity of the pitch axis of the 1031 precision servo gimbal.

[0043] The data receiving and processing unit 200 is connected to the low-altitude surveillance radar 101, the radio spectrum detection device 102, and the electro-optical tracking turntable 103 via physical links. The data receiving and processing unit 200 is configured to perform time synchronization and coordinate transformation processing. The data receiving and processing unit 200 sends a unified time pulse signal and Coordinated Universal Time (UTC) timestamp information to the low-altitude surveillance radar 101, the radio spectrum detection device 102, and the electro-optical tracking turntable 103. The data receiving and processing unit 200 ensures that all data frames output by the sensor acquisition unit 100 are marked with a unified time reference. The installation position coordinates of the low-altitude surveillance radar 101, the radio spectrum detection device 102, and the electro-optical tracking turntable 103 are measured, calibrated, and stored in the data storage and management unit 600 during the system initialization phase. All installation position coordinates are unified to either the geocentric coordinate system or the universal geodetic coordinate system.

[0044] Continue to refer to the appendix Figure 1 The data analysis and situation generation unit 300 establishes a two-way data communication connection with the data receiving and processing unit 200 through a high-speed local area network. The data analysis and situation generation unit 300 is configured to perform tasks such as loading the digital twin geographic base, multi-level data fusion, and identification and inference.

[0045] The data receiving and processing unit 200 is configured to receive raw observation data streams from the sensor acquisition unit 100. The data receiving and processing unit 200 performs protocol parsing and format unpacking on the target point data output by the low-altitude surveillance radar 101, the detection data packets output by the radio spectrum detection device 102, and the photoelectric video stream and photoelectric state vector output by the photoelectric tracking turntable 103. The data receiving and processing unit 200 converts the unpacked heterogeneous data into a unified data frame format within the system. The data receiving and processing unit 200 performs time alignment processing on the multi-source data based on timestamp information to generate a structured spatiotemporal dataset to be processed. Furthermore, the data receiving and processing unit 200 is also used to perform noise filtering and micro-Doppler feature extraction on the radar data, forming motion feature data including velocity vectors and acceleration rates of change.

[0046] The data storage and management unit 600 is configured in the system's storage medium. The data storage and management unit 600 stores three-dimensional geographic information data of the monitored area, sensor configuration parameters, and end-to-end archived data. The three-dimensional geographic information data includes digital elevation model data, oblique photogrammetry three-dimensional mesh model data, and vectorized building outline data.

[0047] The data analysis and situation generation unit 300 reads the three-dimensional geographic information data from the data storage and management unit 600 to construct a digital twin geographic base under a unified coordinate system. The data analysis and situation generation unit 300 receives the real-time photoelectric state vector from the photoelectric tracking turntable 103. The data analysis and situation generation unit 300 drives the virtual camera to perform synchronous movement within the digital twin geographic base based on the azimuth, pitch, and field of view parameters in the photoelectric state vector. The data analysis and situation generation unit 300 utilizes ray tracing algorithms or rasterization rendering pipelines to map the physically detected targets onto the digital twin geographic base, achieving dynamic mapping that blends the virtual and real worlds.

[0048] See attached document Figure 2 This invention provides a comprehensive situational awareness method for countering unmanned aerial vehicles (UAVs), including the steps of acquiring multi-source geographic data, semantic segmentation, spatial registration, and constructing a digital twin geographic base during the system initialization phase. The data analysis and situational awareness generation unit 300 acquires the geospatial basic data of the monitoring area through the data storage and management unit 600. The geospatial basic data includes two-dimensional vector geographic information data and oblique photogrammetry three-dimensional real-scene data. The two-dimensional vector geographic information data contains the coordinate sequences of building base outlines, road edge lines, and functional area boundaries within the monitoring area. The oblique photogrammetry three-dimensional real-scene data includes a set of multi-angle overlapping images collected by low-altitude UAV aerial surveys.

[0049] The data analysis and situation generation unit 300 calls its internal modeling and processing module to perform photogrammetric calculations on the oblique photogrammetric 3D reality data. The modeling and processing module performs aerial triangulation to reconstruct the spatial position and attitude of each image. Based on a multi-view image dense matching algorithm, the modeling and processing module generates a high-density 3D point cloud. Using a meshing algorithm, the modeling and processing module constructs a continuous irregular triangular mesh model from the 3D point cloud. Finally, the modeling and processing module maps image textures onto the surface of the irregular triangular mesh, generating a reality 3D mesh model that includes geometric structure and surface texture.

[0050] The data analysis and situation generation unit 300 performs a unified spatial reference datum transformation on the two-dimensional vector geographic information data and the real-scene three-dimensional mesh model. Since the original geographic data usually uses a geodetic coordinate system (longitude, latitude, elevation), while subsequent ray tracing and geometric collision detection need to be performed in a Cartesian coordinate system, the data analysis and situation generation unit 300 uniformly transforms all spatial data coordinates to a geocentric coordinate system or a station-centered tangential plane coordinate system with the monitoring center as the origin.

[0051] Define the coordinates of any point in the geodetic coordinate system as ,in Latitude of the earth Longitude, The data analysis and situation generation unit 300 uses a coordinate transformation algorithm to convert it into a coordinate vector in the geocentric-ground-fixed coordinate system. .

[0052] The coordinate transformation formula is as follows:

[0053] ;

[0054] ;

[0055] ;

[0056] in: The radius of curvature of the prime and trochanter of the reference ellipsoid is expressed as follows: ; This represents the length of the semi-major axis of the reference ellipsoid (a constant). This represents the first eccentricity (constant) of the reference ellipsoid; This represents the square of the first eccentricity.

[0057] The data analysis and situation generation unit 300 performs spatial registration and fusion of two-dimensional vector data and a real-world three-dimensional mesh model. The data analysis and situation generation unit 300 projects the building base outlines from the two-dimensional vector geographic information data upwards along the vertical direction, forming a virtual columnar space surrounding the buildings. The data analysis and situation generation unit 300 calculates the geometric inclusion relationship between each triangular facet in the real-world three-dimensional mesh model and this virtual columnar space.

[0058] The data analysis and situation generation unit 300 marks a set of triangular faces entirely within the virtual columnar space of a building's base outline as components of that specific building and assigns them unique building identifiers. The data analysis and situation generation unit 300 marks a set of triangular faces located within the road edge line as road surface areas. The data analysis and situation generation unit 300 classifies the remaining set of triangular faces not included in the vector outline as general terrain or vegetation areas. Through these steps, the data analysis and situation generation unit 300 completes the construction of a structured digital twin scene with semantic objects from an unstructured 3D mesh, and stores the processed data in the data storage and management unit 600 as the geographic basis for subsequent multi-source data fusion and situation display.

[0059] After completing the geometric fusion of multi-source geographic data, the data analysis and situation generation unit 300 further defines the physical properties of the surface attributes of the real-scene 3D mesh model stored in the data storage and management unit 600. The data analysis and situation generation unit 300 calls its internal semantic segmentation module to perform pixel-level classification of the texture maps corresponding to the 3D mesh model.

[0060] The data analysis and situation generation unit 300 employs a semantic segmentation algorithm based on a deep convolutional neural network to divide pixel regions in the texture map into different material categories. These material categories include concrete, asphalt, glass, metal, vegetation, water, and exposed soil. Based on the texture map segmentation results, the data analysis and situation generation unit 300 logically divides the triangular facet set in the 3D mesh model into sub-mesh objects with a single material attribute. For mesh models identified as buildings, the data analysis and situation generation unit 300 further subdivides them into wall sub-mesh, window glass sub-mesh, and roof sub-mesh.

[0061] The data analysis and situation generation unit 300 traverses each triangular facet in the real-world 3D mesh model, retrieving corresponding physical parameters from the physical material attribute library based on the material category of the facet. The data analysis and situation generation unit 300 generates a feature vector containing physical attributes for each triangular facet, supporting subsequent feature layer fusion and infrared simulation.

[0062] Definition of the first The physical property feature vector of each triangular facet as follows:

[0063] ;

[0064] in: This represents the average diffuse reflectance of the triangular facet in the visible light band, with a value ranging from 0 to 1. This represents the thermal emissivity of the triangular facet in the long-wave infrared band, with a value ranging from 0 to 1. This represents the specular reflection coefficient of the triangular facet, used to describe the smoothness of the surface, and its value ranges from 0 to 1. This represents the heat capacity coefficient of the triangular facet, used for subsequent thermodynamic temperature field simulations.

[0065] The data analysis and situation generation unit 300 also needs to calculate and store the geometric normal vector of each triangular facet for subsequent illumination calculations and signal reflection path deduction. The data analysis and situation generation unit 300 extracts the... The coordinates of the three vertices of the triangular facet are respectively The data analysis and situation generation unit 300 calculates the unit normal vector of the triangular facet according to the right-hand rule. .

[0066] Unit normal vector The calculation formula is as follows:

[0067] ;

[0068] in: Represents the cross product operation of vectors; The Euclidean norm (modulus) of a vector.

[0069] The data analysis and situation generation unit 300 executes a manual or automatic marking procedure for specific heat sources. For known fixed heat source facilities within the airport area (including air conditioning outdoor units, generator exhaust vents, and substation heat sinks), the data analysis and situation generation unit 300 marks the corresponding triangular facets with physical attribute feature vectors. A specific heat source identifier is set in the system, and a preset base temperature rise value is assigned.

[0070] The data analysis and situation generation unit 300 will include physical attribute feature vectors. and unit normal vector The enhanced 3D mesh model is then re-stored into the data storage and management unit 600, forming a digital twin geographic foundation with physical interactive computing capabilities. Through this step, the digital twin scene transforms from a simple visual display model into a computational model capable of supporting multi-source data fusion and analysis.

[0071] The data analysis and situation generation unit 300 is configured to construct a dynamically evolving environmental parameter field to drive lighting rendering and thermodynamic simulation in the digital twin scene. The data analysis and situation generation unit 300 connects to a meteorological monitoring station via a network interface to acquire real-time meteorological data for the monitored area. The meteorological data includes ambient air temperature, relative humidity, and total horizontal radiation intensity.

[0072] The data analysis and situation generation unit 300 executes the solar ephemeris calculation program. Based on the current Coordinated Universal Time (UTC) timestamp and the center latitude and longitude coordinates of the monitored area provided by the data receiving and processing unit 200, the data analysis and situation generation unit 300 calculates the position parameters of the sun in the local celestial coordinate system at the current moment. The position parameters include the solar altitude angle and the solar azimuth angle. Based on the calculated position parameters, the data analysis and situation generation unit 300 constructs a normalized solar incident light vector in the local tangent plane coordinate system of the digital twin scene. .

[0073] Sun incident light vector The calculation formula is as follows:

[0074] ;

[0075] in: It represents the solar altitude angle, which is the angle between the sun's rays and the horizontal plane; It represents the solar azimuth angle, which is the angle between the projection of the sun's rays onto the ground plane and the due north direction (clockwise is positive); the three components of the vector correspond to the component values ​​of the east, north, and sky axes, respectively.

[0076] The data analysis and situation generation unit 300 performs real-time simulation updates of the ground surface temperature field. The data analysis and situation generation unit 300 iterates through each triangular facet stored in the spatiotemporal database 202. The data analysis and situation generation unit 300 then updates the data based on the current ambient air temperature and the solar incident light vector. The total horizontal radiation intensity and the physical property characteristic vector of the triangular facet. Calculate the theoretical surface temperature of the triangular facet at the current moment.

[0077] Definition of the first The theoretical surface temperature of a triangular facet as follows:

[0078] ;

[0079] in: Indicates the ambient air temperature at the current moment; Indicates the first The heat capacity coefficient of each triangular facet (derived from the physical property feature vector) ); This represents the total horizontal radiation intensity at the current moment; Indicates the first The unit normal vector of a triangular facet; Represents the dot product operation of vectors; This indicates the maximum value operation, which ensures that the direct sunlight radiation temperature increase is calculated only if the triangular facet faces the sun; the value is zero on the shaded side. Indicates the first Temperature rise value of the internal fixed heat source of each triangular facet (this value is zero if the facet is marked as a non-heat source facility).

[0080] The data analysis and situation generation unit 300 calculates the theoretical surface temperature based on the Stefan-Boltzmann law. The data is converted to an infrared radiation intensity value. The data analysis and situation generation unit 300 uses this radiation intensity value to update the rendered texture data of each triangle in the digital twin scene. For the visible light channel, the data analysis and situation generation unit 300 uses the solar incident light vector... and the diffuse reflectance of triangular facets The system calculates the reflected brightness. The data analysis and situation generation unit 300 caches a dynamic texture layer containing real-time infrared radiation intensity and visible light brightness into the video memory of the digital twin rendering engine 203, serving as the rendering reference for subsequently generating the expected background image. Through this step, the system completes the real-time mapping of physical environment parameters to digital twin visual features.

[0081] See attached document Figure 3 This invention provides a comprehensive situational awareness method for countering unmanned aerial vehicles (UAVs), including steps of spatiotemporal alignment, preprocessing, and feature extraction of multi-source heterogeneous sensor data. The data receiving and processing unit 200 receives the raw data stream uploaded by the sensor acquisition unit 100 in real time. The data receiving and processing unit 200 establishes a unified time axis based on the system clock signal. The data receiving and processing unit 200 performs timestamp alignment on the data frames from the low-altitude surveillance radar 101, the radio spectrum detection device 102, and the electro-optical tracking turntable 103. For data sources with inconsistent sampling rates, the data receiving and processing unit 200 uses zero-order hold interpolation to align low-frequency data to the sampling time of high-frequency data.

[0082] The data receiving and processing unit 200 performs spatial coordinate transformation on the radar spot data. The data receiving and processing unit 200 reads the installation position coordinates of the low-altitude surveillance radar 101 stored in the data storage and management unit 600. The data receiving and processing unit 200 receives the target slant range output by the radar. Azimuth and pitch angle The data receiving and processing unit 200 constructs a rotation transformation matrix from the radar local coordinate system to the geocentric coordinate system. The data receiving and processing unit 200 converts the target point trace in the radar polar coordinate system into a three-dimensional position vector under a unified spatiotemporal reference. .

[0083] 3D position vector The calculation formula is as follows:

[0084] ;

[0085] in: This represents the origin position vector of the low-altitude surveillance radar 101 in the geocentric coordinate system. This represents the rotation matrix from the radar station's tangential plane coordinate system to the Earth-centered, Earth-fixed coordinate system; Represents matrix multiplication operations; matrix column vectors This indicates the relative position of the target in the radar's local Cartesian coordinate system.

[0086] The data receiving and processing unit 200 establishes a pinhole camera projection model of the photoelectric imaging system. The data receiving and processing unit 200 reads the real-time photoelectric state vector output by the photoelectric tracking turntable 103. The data receiving and processing unit 200 receives data based on the azimuth angle of the precision servo gimbal 1031. Pitch angle and the installation location of the visible light camera 1032 or the infrared thermal imager 1033. Construct the viewpoint transformation matrix from the world coordinate system to the camera coordinate system. The data receiving and processing unit 200 receives and processes data based on the current lens focal length. And sensor pixel size, construct camera intrinsic parameter matrix The data receiving and processing unit 200 uses this projection model to describe the mapping relationship between three-dimensional spatial points and two-dimensional image pixel planes.

[0087] The mapping formula for the pinhole camera projection model is as follows:

[0088] ;

[0089] in: Represents the pixel coordinate vector on the image plane ; This represents the depth value of the target point in the camera coordinate system; This represents the camera intrinsic parameter matrix, which includes focal length and principal point offset. This represents the viewpoint transformation matrix (extrinsic parameter matrix) that includes rotation and translation components. Represents the homogeneous coordinate vector of a point in the geocentric coordinate system. .

[0090] The data receiving and processing unit 200 performs denoising and enhancement preprocessing on the acquired raw image data. The data receiving and processing unit 200 applies a Gaussian smoothing filter to reduce image noise and suppress sensor thermal noise. The data receiving and processing unit 200 applies a histogram equalization algorithm to enhance image contrast. The data receiving and processing unit 200 then transmits the processed structured spatiotemporal dataset to the data analysis and situation generation unit 300.

[0091] See attached document Figure 4 This invention provides a comprehensive situational awareness method for countering unmanned aerial vehicles (UAVs), comprising multi-level data fusion steps including an execution data layer, a feature layer, and a decision layer. The data analysis and situational awareness generation unit 300 synchronously receives radar motion characteristics, radio frequency characteristics, and photoelectric image characteristics from the data receiving and processing unit 200.

[0092] The data analysis and situation generation unit 300 first performs data layer fusion. The data analysis and situation generation unit 300 sets a spatial correlation gate and performs cluster analysis on time-aligned data from different sensors. For radar traces, radio location points, and visual target points falling within the same correlation gate, the data analysis and situation generation unit 300 determines that they originate from the same physical entity and groups them into a target group to be fused.

[0093] The data analysis and situation generation unit 300 then performs feature layer fusion and decision layer fusion. The data analysis and situation generation unit 300 reads real-time environmental parameters uploaded by the meteorological monitoring station, including visibility distance, rainfall intensity, and ambient illuminance. Based on these environmental parameters, the data analysis and situation generation unit 300 dynamically adjusts the measurement noise covariance matrix of each sensor. At the decision layer, the data analysis and situation generation unit 300 combines the micro-Doppler features of radar (to distinguish rotors from birds), the radio frequency fingerprint (to identify communication protocols), and the texture features of photoelectric images (morphological classification), employing a Bayesian inference network or support vector machine (SVM) to classify and identify targets.

[0094] Definition of the first The adaptive measurement noise covariance matrix of each sensor under current environmental conditions as follows:

[0095] ;

[0096] in: Indicates the first The noise covariance matrix of a sensor under a benchmark measurement in an ideal laboratory environment; This represents the total number of environmental factors affecting this type of sensor; Indicates the first Normalized intensity values ​​of environmental factors (e.g., normalized haze concentration or rain attenuation coefficient); Indicates the first Environmental factors on the first Interference sensitivity coefficient of each sensor.

[0097] The data analysis and situation generation unit 300 employs an optimal weighted fusion algorithm based on an information filtering architecture to calculate the target's comprehensive state vector. The comprehensive state vector includes the target's position, velocity, and acceleration components in three-dimensional space. The data analysis and situation generation unit 300 uses the inverse matrix (i.e., the information matrix) of the adaptive measurement noise covariance matrix of each sensor as weighting coefficients to perform a linear weighted combination of the observation states from each sensor.

[0098] Define the target integrated state vector after fusion. The calculation formula is as follows:

[0099] ;

[0100] in: This indicates the total number of sensors involved in the fusion process; Indicates the first The information matrix of each sensor, that is, the inverse matrix of the adaptive measurement noise covariance matrix; Indicates the first The observation state vector output by each sensor (already converted to a common coordinate system).

[0101] The data analysis and situation generation unit 300 will calculate the comprehensive state vector. The final target is three-dimensional airspace situational data. The data analysis and situational generation unit 300 sends this situational data to the integrated situational 2D / 3D display unit 400. The integrated situational 2D / 3D display unit 400 plots the fused target trajectory on a digital 3D map and displays the corresponding target symbol according to the category attribute in the fusion result. When the fusion result confirms that the target is an intruding UAV, the integrated situational 2D / 3D display unit 400 automatically generates an alarm signal.

[0102] See attached document Figure 5 This invention provides a comprehensive situational awareness method for countering unmanned aerial vehicles (UAVs), including a step of performing a two-dimensional / three-dimensional integrated situational awareness presentation. The comprehensive situational awareness two-dimensional / three-dimensional display unit 400 is equipped with a two-dimensional / three-dimensional view linkage controller 400a, a two-dimensional electronic map rendering engine 400b, and a three-dimensional digital twin rendering engine 400c. The comprehensive situational awareness two-dimensional / three-dimensional display unit 400 receives the target comprehensive state vector output by the multi-source fusion solver 205. .

[0103] The 2D electronic map rendering engine 400b loads locally stored vector geographic information tile data to construct a 2D orthophoto view. Based on the latitude and longitude coordinates in the target's integrated state vector, the 2D electronic map rendering engine 400b draws the target's tactical symbols, historical track lines, and velocity vector arrows on the 2D view layer. The 2D electronic map rendering engine 400b uses different colors to fill the tactical symbols according to the target's threat level attributes.

[0104] The 3D digital twin rendering engine 400c loads a physically-based 3D mesh model with real-world properties from the data storage and management unit 600, constructing a 3D perspective projection view. The 3D digital twin rendering engine 400c establishes a virtual lighting environment in the 3D scene, with parameters synchronized with the physical environment parameter field. Based on the 3D spatial coordinates in the target's integrated state vector, the 3D digital twin rendering engine 400c renders the target's 3D model and its vertical projection lines relative to the ground in the 3D scene.

[0105] The 2D / 3D view linkage controller 400a performs view synchronization calculations. When an operator performs panning, zooming, or target locking operations in the view generated by the 2D electronic map rendering engine 400b, the 2D / 3D view linkage controller 400a captures the geographic coordinates of the center point of the current 2D view in real time. and the current map scale levels The 2D / 3D view linkage controller 400a calculates the spatial pose parameters of the virtual camera in the 3D digital twin rendering engine 400c based on these parameters.

[0106] Define the spatial position vector of the virtual camera The calculation formula is as follows:

[0107] ;

[0108] in: This represents the three-dimensional coordinate vector of the center point of a two-dimensional view in the geocentric coordinate system. This represents the line-of-sight function, used to determine the line-of-sight level based on the scale of a two-dimensional map. Calculate the Euclidean distance between the virtual camera and the observation center point; Indicates the preset pitch angle of the virtual camera; The preset azimuth angle of the virtual camera is represented; the column vectors of the matrix represent the unit direction vectors at a given azimuth and elevation angle.

[0109] The 400a two- or three-dimensional view linkage controller will calculate the... and by The determined viewing point is sent to the 3D digital twin rendering engine 400c. Based on this parameter, the 3D digital twin rendering engine 400c immediately updates the view matrix of the virtual camera, achieving frame-level synchronization between the 3D scene viewport and the 2D map viewport.

[0110] The integrated situation 2D / 3D display unit 400 performs field-of-view analysis and visibility calculation. Utilizing geometric data from the 3D digital twin scene, the integrated situation 2D / 3D display unit 400 calculates the line-of-view visibility from the current radar station or electro-optical turntable position to the target position. If the line of sight is obstructed by buildings or terrain, the integrated situation 2D / 3D display unit 400 draws the obstruction segment as a dashed line in the 3D view and displays an obstruction status indicator next to the target symbol in the 2D view. The integrated situation 2D / 3D display unit 400 displays the 2D global situation map and the 3D local detail map on the same screen, providing commanders with a comprehensive decision-making basis combining macro and micro perspectives.

[0111] See attached document Figure 6 This invention provides a comprehensive situational awareness method for countering unmanned aerial vehicles (UAVs), including calibration feedback and closed-loop control steps. The comprehensive situational awareness 2D / 3D display unit 400 is equipped with a human-machine interface feedback interface 206d, which receives confirmation, correction, or rejection commands from operators regarding the current alarm target. When an operator marks a "suspected UAV target" automatically identified by the system as a "false alarm" or corrects its target type label on the comprehensive situational awareness 2D / 3D display unit 400 via an input device, the monitoring object calibration and identification unit 500 generates a feedback data packet containing a unique target identifier, a timestamp of the alarm occurrence time, and the corrected target attribute label.

[0112] The monitoring target identification and labeling unit 500 sends the feedback data packet to the data analysis and situation generation unit 300 and the data storage and management unit 600. Based on the manual identification results in the feedback data packet, the data analysis and situation generation unit 300 updates the current situation display in real time, removing false alarm targets or correcting target icons.

[0113] Meanwhile, the data storage and management unit 600 archives the entire data chain. The data storage and management unit 600 establishes a spatiotemporal index database for classifying and storing raw monitoring data, fusion analysis results, and manually calibrated tags. The system also includes a situational data playback and control module, which retrieves historical data by time period or target ID and enables synchronous playback and speed control of historical situations in the integrated situational 2D / 3D display unit 400. Through a closed-loop mechanism of physical parameter correction, manual calibration feedback, and historical data playback, the system ensures continuous consistency between the digital twin scenario and the physical reality scenario, as well as the accuracy of situational awareness.

Claims

1. A comprehensive anti-drone situational awareness system based on multi-source data fusion, characterized in that, include: The system includes a sensor acquisition unit, a data receiving and processing unit, a data analysis and situation generation unit, a comprehensive situation 2D / 3D display unit, a monitoring object calibration and identification unit, and a data storage and management unit. The sensor acquisition unit is used to integrate a heterogeneous sensor group deployed in the monitoring area to acquire raw target data; The data receiving and processing unit is connected to the sensor acquisition unit and is used to preprocess the target raw data and output a structured spatiotemporal observation dataset. The data analysis and situation generation unit is connected to the data receiving and processing unit. It is used to load the digital twin geographic base of the monitoring area and perform multi-level data fusion and identification inference based on the spatiotemporal observation dataset to generate three-dimensional airspace situation data. The integrated situation 2D / 3D display unit is connected to the data analysis and situation generation unit and is used to visualize the spatial situation data. The monitoring object calibration and identification unit is connected to the comprehensive situation two-dimensional / three-dimensional display unit and the data analysis and situation generation unit, and is used to provide a human-computer interaction interface to correct target attributes; The data storage and management unit connects the above-mentioned units and is used for archiving and closed-loop management of data across the entire chain.

2. The anti-drone integrated situational awareness system based on multi-source data fusion according to claim 1, characterized in that, The sensor acquisition unit includes a low-altitude surveillance radar, an optoelectronic tracking device, and a radio spectrum analyzer. The low-altitude surveillance radar is used to output point data containing target range, azimuth, altitude, and radial velocity. The photoelectric tracking device is used to acquire visible light or infrared video streams and output the optical axis pointing angle of the device; The radio spectrum analyzer is used to capture the target's communication signals and output the signal frequency and spectrum characteristics.

3. The anti-drone integrated situational awareness system based on multi-source data fusion according to claim 1, characterized in that, The data receiving and processing unit is equipped with a standard communication interface and a protocol parsing module for real-time access to the target raw data. The data receiving and processing unit is also used to perform time synchronization and coordinate transformation processing, unifying the measurement data from different sensors to the Coordinated Universal Time (UTC) time base and the WGS-84 geographic coordinate system; The data receiving and processing unit is also used to perform noise filtering and micro-Doppler feature extraction on radar data to form motion feature data containing velocity vectors and acceleration change rates.

4. The anti-drone integrated situational awareness system based on multi-source data fusion according to claim 1, characterized in that, The data analysis and situation generation unit includes a geographic information preprocessing module, which is used to load basic GIS vector data and oblique photogrammetry 3D model, and construct the digital twin geographic base through spatial registration. The data analysis and situation generation unit is used to map the processed sensor data to the digital twin geographic base, realizing the dynamic mapping from physical detection targets to digital twins.

5. The anti-drone integrated situational awareness system based on multi-source data fusion according to claim 1, characterized in that, The data fusion performed by the data analysis and situation generation unit includes data layer fusion, feature layer fusion, and decision layer fusion. At the data layer, radar traces and photoelectric images are spatiotemporally aligned and correlated. At the feature layer, the vector motion features of radar, the radio frequency fingerprint features of radio waves, and the texture features of optoelectronic images are fused. At the decision-making level, rule-based or machine learning algorithms are used to classify and identify targets, distinguishing between drones, biomimetic drones, and bird targets, and outputting a unified situational awareness object that includes target ID, type, trajectory, and threat level.

6. The anti-drone integrated situational awareness system based on multi-source data fusion according to claim 1, characterized in that, The integrated situation 2D / 3D display unit is used to drive the synchronous display of the 2D electronic map and the 3D electronic sand table; In two-dimensional electronic maps, the geographical location and historical flight path of a target are displayed using symbolic icons; In the three-dimensional electronic sand table, the real geographical environment is rendered based on the oblique photogrammetry three-dimensional model, and the target is embedded in the geographical environment in the form of a three-dimensional model, intuitively displaying the target's height, posture and relative relationship with surrounding buildings.

7. The anti-drone integrated situational awareness system based on multi-source data fusion according to claim 6, characterized in that, The integrated situation 2D / 3D display unit supports a view linkage mechanism. When a specific target is selected in either the 2D electronic map or the 3D electronic sand table, the other view will automatically highlight the target or switch the viewpoint to the center.

8. The anti-drone integrated situational awareness system based on multi-source data fusion according to claim 1, characterized in that, The monitoring object calibration and identification unit is used to respond to the operator's calibration instructions for suspicious targets and manually correct the target type or threat level labels in the airspace situation data; The corrected tag data is fed back to the data analysis and situation generation unit and the data storage and management unit in real time to update the current situation display and database records, forming an optimized closed loop of human-machine collaboration.

9. The anti-drone integrated situational awareness system based on multi-source data fusion according to claim 1, characterized in that, The data storage and management unit establishes a spatiotemporal index database for classifying and storing raw monitoring data, fusion analysis results, and manually calibrated labels; The system also includes a situation data playback and control module, which is used to retrieve historical data by time period or target ID, and realize synchronous playback and speed control of historical situations in the integrated situation two-dimensional / three-dimensional display unit.

10. A comprehensive situational awareness method for countering unmanned aerial vehicles (UAVs) based on multi-source data fusion, characterized in that, The anti-drone integrated situational awareness system based on multi-source data fusion as described in any one of claims 1-9 includes the following steps: A digital twin geographic base under a unified coordinate system is constructed using GIS vector data and oblique photogrammetry 3D models; The sensor acquisition unit acquires multi-source heterogeneous data from radar, photoelectric and radio sources in real time, and uses the data receiving and processing unit to perform protocol parsing, time synchronization and coordinate unification, and outputs a structured spatiotemporal dataset. The structured spatiotemporal dataset is fused in multiple levels using a data analysis and situation generation unit. Target types are identified by combining radar motion characteristics, radio frequency characteristics, and photoelectric texture characteristics, generating an airspace situation object containing three-dimensional position and attribute information. The airspace situation object is mapped to the digital twin geographic base, driving the integrated situation two-dimensional / three-dimensional display unit to simultaneously present the target's dynamic trajectory in the two-dimensional map and three-dimensional sand table; The monitoring object identification and labeling unit receives manual review instructions, corrects the target classification labels, and updates the correction results to the data storage and management unit in real time, thus completing the closed-loop control of situational awareness.

Images