Low-altitude economic airspace passive safety early warning method and system

Abstract

The invention relates to the technical field of urban low-altitude safety monitoring, and discloses a low-altitude economic airspace passive safety early warning method and system. The method comprises the following steps: collecting multi-modal observation data streams of a target at different moments; generating multi-modal frame data with a unified timestamp; establishing a spatial confidence distribution map; extracting fusion discrimination features; executing multi-channel small target detection; continuously tracking the preliminary target; and introducing a constant false alarm rate framework to set a dynamic confidence threshold to execute alarm judgment. Compared with the prior art, the technical problem that high-confidence all-weather passive monitoring cannot be realized in urban core areas with dense buildings, serious signal shielding and easy GPS failure due to dependence on single vision or radar information in the prior art is solved. According to the method, the space confidence distribution diagram and the dynamic constant false alarm mechanism based on the shielding diagram are constructed by fusing the multi-modal sensing data, so that the safety early warning of the low-altitude small aircraft is realized, and the passive safety guarantee capability of the low-altitude economic airspace is improved.

Description

Technical Field

[0001] This invention relates to the field of urban low-altitude safety monitoring technology, and in particular to an early warning method and system for passive safety in low-altitude economic airspace. Background Technology

[0002] Currently, urban core areas are increasingly becoming major areas for illegal low-altitude flight activities (such as unauthorized drone flights). Especially in complex environments with towering buildings and significant electromagnetic interference, traditional active radar and single-modal vision methods have obvious shortcomings in terms of accuracy and real-time performance in small target recognition. For example, video detection models are prone to failure in environments with severe obstruction, large changes in lighting, or strong reflections; infrared sensing is easily interfered with by high-temperature building backgrounds; and RF signal detection is often affected by frequency drift, multipath interference, and low-power spoofing signals, resulting in high false alarm rates and delayed response.

[0003] Existing technologies often rely on single-modality or single-point sensors for target monitoring, lacking multimodal collaborative perception strategies for the complex spatiotemporal characteristics of urban core areas. They cannot achieve cross-sensor time alignment, occlusion compensation, and stable identification of targets with low signal-to-noise ratios. Especially in the presence of highly dynamic occluded objects (such as elevated vehicles, pedestrians, etc.) or short-term signal loss, existing technologies cannot fully meet the needs of real-time response and continuous monitoring for low-altitude safety.

[0004] Therefore, there is an urgent need for a passive safety early warning method for low-altitude economic airspace that can still identify and continuously track low-altitude micro-targets in complex urban environments with multiple sources of obstruction, electromagnetic interference, and low signal-to-noise characteristics of small targets, so as to improve the safety management level and automatic response capability of urban low-altitude economic airspace. Summary of the Invention

[0005] To address the aforementioned technical shortcomings, the purpose of this invention is to propose a passive safety early warning method for low-altitude economic airspace. This method aims to solve the technical problem that existing technologies rely on single visual or radar information, especially in densely built-up urban core areas with severe signal obstruction and GPS malfunction, making it impossible to achieve high-confidence, all-weather passive monitoring.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides an early warning method for passive safety in low-altitude economic airspace. The early warning methods for passive safety in low-altitude economic airspace include: Step S10: By deploying a sensor array in the core urban area, real-time data is acquired based on the sensor array at any given time. The observation data stream is processed; clock synchronization is performed on the observation data stream using a combination of GNSS terminal timing and NTP synchronization protocol, outputting a unified timestamp; based on the unified timestamp, cross-correlation analysis is used to perform time deviation compensation processing on the observation data stream, outputting multimodal frame data. ; Step S20: Obtain GIS building model data and sensor view parameters, and construct an occlusion map based on the GIS building model data and sensor view parameters. Based on occlusion map and multimodal frame data A spatial confidence distribution map is established using a priori perception capability estimation mechanism. And based on the spatial confidence distribution map Determine the priority modal frame set ; Step S30: Set the priority modal frames Input to a pre-defined multimodal target detection network Multimodal target detection network Output fusion discriminative features Based on fusion discriminative features The YOLOv7 vision detection library in Python and the pre-defined low-altitude signal standard function library are used to perform a multi-channel small target detection task, and a preliminary target set is output. ; Step S40: Based on the preliminary target set Establish the target state vector And based on the target state vector Predicting the i-th target at the next time step using the extended Kalman filter method +1 target state The target state is tracked using an RF drift compensation tracking mechanism. Perform continuous predictive tracking and output the trajectory estimation result of the i-th target; Step S50: Based on the trajectory estimation results, a confidence verification and constant false alarm rate fusion mechanism is used to trigger an effective detection alarm event.

[0007] Preferably, in step S10, the sensor array includes a radar sensor, an infrared thermal imaging sensor, a visible light sensor, and an RF receiver.

[0008] Preferably, in step S10, the observation data stream includes millimeter-wave radar echo signals. Thermal infrared image frame sequence RGB video image frames Channel status information of RF spectrum .

[0009] Preferably, in step S20, GIS building model data and sensor view parameters are acquired, and an occlusion map is constructed based on the GIS building model data and sensor view parameters. Based on occlusion map and multimodal frame data A spatial confidence distribution map is established using a priori perception capability estimation mechanism. And based on the spatial confidence distribution map Determine the priority modal frame set The steps specifically include: Step S201: Obtain GIS building model data and sensor view parameters, and construct an occlusion map based on the GIS building model data and sensor view parameters using the view volume projection method. ; Occlusion image Used to characterize whether each point in space is within the visible area; Step S202: Introduce a perception intensity mapping function, which is used to estimate the perception probability of different modal data at different points in three-dimensional space; based on the occlusion map... and multimodal frame data By combining the perception intensity mapping function, the comprehensive observability score at each three-dimensional spatial point is calculated; Step S203: When the overall observability score is greater than the preset perception score threshold, the three-dimensional spatial point is determined to be a recognizable region, and the priority modality frame set is finally output. .

[0010] Preferably, in step S30, the multimodal target detection network Includes a multimodal input normalization layer for receiving a priority modal frame set. Furthermore, by performing size resampling, dynamic range normalization, and channel alignment operations, the priority modal frame set is... The process involves unifying the tensor representation to a unified form, forming a fused input tensor; a modal feature extraction layer, which encodes the features of different modal tensors in the fused input tensor to obtain modal feature maps; a modal attention fusion layer, which performs weighted fusion based on the modal feature maps using an attention mechanism to generate a fused feature map; and a multi-scale detection decoding layer, which performs spatial scale decoding based on the fused feature map to output fused discriminative features. .

[0011] Preferably, in step S40, an RF drift compensation tracking mechanism is used to track the target state. The steps for performing continuous predictive tracking and outputting the trajectory estimation result of the i-th target specifically include: acquiring the target state from the RF receiver. The corresponding Channel State Change Feature (CSI) is used to extract the center drift rate of the CSI within a preset sliding window, and the dynamic fusion coefficient is set based on the center drift rate. , dynamic fusion coefficient Applied to target state Perform a weighted combination and output the trajectory estimation result of the i-th target.

[0012] Preferably, in step S50, the step of triggering an effective detection alarm event based on the trajectory estimation result using a confidence verification and constant false alarm rate (CFAR) fusion mechanism specifically includes: introducing a constant false alarm rate (CFAR) framework and setting a dynamic confidence threshold based on the CFAR framework. Obtain the spatial confidence score corresponding to the trajectory estimation result. When the spatial confidence score Greater than or equal to the dynamic confidence threshold When a valid alarm event is detected, it is determined that there is a valid alarm event. Valid alarm events include illegal low-altitude flight alarm events, static camouflage target drift and sudden change alarm events, target reappearance after continuous obscuration alarm events, and electromagnetic disturbance source alarm events; and the valid alarm events are reported to the regional monitoring platform.

[0013] This invention also provides an early warning system for passive safety in low-altitude economic airspace, comprising: The spatiotemporal synchronization and data preprocessing module is used to acquire real-time data based on sensor arrays deployed in the city's core area. The observation data stream is processed; clock synchronization is performed on the observation data stream using a combination of GNSS terminal timing and NTP synchronization protocol, outputting a unified timestamp; based on the unified timestamp, cross-correlation analysis is used to perform time deviation compensation processing on the observation data stream, outputting multimodal frame data. ; The view modeling and perception confidence distribution construction module is used to acquire GIS building model data and sensor view parameters, and to construct an occlusion map based on the GIS building model data and sensor view parameters. Based on occlusion map and multimodal frame data A spatial confidence distribution map is established using a priori perception capability estimation mechanism. And based on the spatial confidence distribution map Determine the priority modal frame set ; The multimodal target detection and feature fusion module is used to combine the priority modal frame set. Input to a pre-defined multimodal target detection network Multimodal target detection network Output fusion discriminative features Based on fusion discriminative features The YOLOv7 vision detection library in Python and the pre-defined low-altitude signal standard function library are used to perform a multi-channel small target detection task, and a preliminary target set is output. ; The target tracking and RF drift compensation prediction module is used to predict targets based on a preliminary target set. Establish the target state vector And based on the target state vector Predicting the i-th target at the next time step using the extended Kalman filter method +1 target state The target state is tracked using an RF drift compensation tracking mechanism. Perform continuous predictive tracking and output the trajectory estimation result of the i-th target; The trajectory verification and alarm triggering module is used to trigger effective detection alarm events based on the trajectory estimation results using a confidence verification and constant false alarm rate fusion mechanism.

[0014] The present invention also provides a low-altitude economic airspace passive safety early warning device, comprising: a memory, a processor, and a low-altitude economic airspace passive safety early warning program stored in the memory and executable on the processor. When the low-altitude economic airspace passive safety early warning program is executed by the processor, a low-altitude economic airspace passive safety early warning method is implemented.

[0015] The present invention also provides a computer program product, including a low-altitude economic airspace passive safety early warning program, which, when executed by a processor, implements the low-altitude economic airspace passive safety early warning method.

[0016] The beneficial effects of this invention are as follows: by constructing a spatial confidence distribution map based on GIS occlusion map and sensor field parameters, and combining a priority mode selection and fusion discrimination feature extraction mechanism, the continuous detection capability of low-altitude small targets (such as drones and illegal signal sources) in complex urban building environments is significantly improved, and the passive perception coverage and recognition accuracy in high occlusion and low signal-to-noise ratio scenarios are effectively enhanced.

[0017] A multi-target trajectory estimation mechanism combining extended Kalman filtering and RF drift compensation was introduced, and a constant false alarm rate dynamic threshold judgment strategy was integrated to effectively reduce the false alarm rate and missed detection rate under urban interference background, achieve robust tracking of weak moving targets and high confidence triggering alarm, and improve the safety management level of urban low-altitude economic airspace. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the first embodiment of a passive safety early warning method for low-altitude economic airspace according to the present invention.

[0020] Figure 2 This is a schematic diagram of an early warning method for passive safety in low-altitude economic airspace according to the present invention. Detailed Implementation

[0021] 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 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.

[0022] Example 1: As Figure 1 The diagram shown is a flowchart of the first embodiment of the early warning method for passive safety in low-altitude economic airspace of the present invention, which presents the first embodiment of the early warning method for passive safety in low-altitude economic airspace of the present invention.

[0023] In the first embodiment, the early warning method for passive safety in low-altitude economic airspace includes: Step S10: By deploying a sensor array in the core urban area, real-time data is acquired based on the sensor array at any given time. The observation data stream is processed; clock synchronization is performed on the observation data stream using a combination of GNSS terminal timing and NTP synchronization protocol, outputting a unified timestamp; based on the unified timestamp, cross-correlation analysis is used to perform time deviation compensation processing on the observation data stream, outputting multimodal frame data. ; It should be noted that "observation data stream" refers to the raw time-series signal data collected by various sensors deployed in the core urban area, such as acoustic arrays, optical cameras, and low-frequency radar. This includes different modalities, video frames, acoustic time-domain signals, electromagnetic spectrum samples, etc., and their original sampled values ​​at different times. "Unified timestamp" refers to the standardized time reference assigned to all observation data after collaborative time synchronization at the edge gateway between the GNSS (Global Navigation Satellite System) terminal and the NTP (Network Time Protocol) service, ensuring accurate registration and synchronous calculation between different modalities.

[0024] Understandably, by introducing a hybrid clock calibration mechanism of GNSS and NTP during the acquisition phase and combining it with cross-correlation functions for inter-frame timing fine-tuning, spatial reconstruction errors or missed event detection problems caused by asynchronous acquisition of multimodal data are effectively avoided, providing timing alignment assurance for subsequent low-altitude target detection tasks based on frame coordination.

[0025] It should be understood that, compared to traditional time synchronization methods using a single reference source (such as using only NTP or internal crystal oscillators), this step can still achieve multi-source data time synchronization accuracy in environments with dense urban high-rise buildings and severe network jitter, significantly improving modal fusion efficiency and error suppression capabilities.

[0026] Step S20: Obtain GIS building model data and sensor view parameters, and construct an occlusion map based on the GIS building model data and sensor view parameters. Based on occlusion map and multimodal frame data A spatial confidence distribution map is established using a priori perception capability estimation mechanism. And based on the spatial confidence distribution map Determine the priority modal frame set ; It should be noted that the "occlusion map" in this step refers to the occlusion area layout derived from GIS 3D building model data and the actual deployment location and field of view parameters of each sensor, used to identify areas in space where each modality may have a perception blind spot; the "prior perception capability estimation mechanism" includes a perception scoring function model constructed based on building occlusion density, historical sensor perception coverage, and data frame integrity; the "spatial confidence distribution map" is a 3D perception confidence heat map formed by mapping the above estimates onto the urban spatial grid, used to select the priority modality frame set with higher information richness.

[0027] Understandably, this step, by constructing a spatial occlusion constraint model, explicitly removes modal frames with weak perception capabilities or severe interference, effectively improving the data utilization efficiency and information density in the subsequent multimodal fusion detection stage.

[0028] It should be understood that traditional object detection methods often treat all modal signals equally, ignoring the impact of high occlusion and multipath distortion in urban core areas on the severe degradation of some modalities. This method optimizes the mechanism of "prioritizing the scheduling of sensing resources" through confidence distribution modeling and modality selection.

[0029] Step S30: Set the priority modal frames Input to a pre-defined multimodal target detection network Multimodal target detection network Output fusion discriminative features Based on fusion discriminative features The YOLOv7 vision detection library in Python and the pre-defined low-altitude signal standard function library are used to perform a multi-channel small target detection task, and a preliminary target set is output. ; It should be noted that "multimodal object detection network" refers to a deep fusion neural network structure containing multiple input modality channels, typically composed of the following layers: input modality encoding layer (for unifying the feature scale of each modality), feature fusion cross-attention layer (for guiding intermodal collaborative attention to key regions), channel selection gating layer (for adaptively enhancing salient modalities), and fusion classification output layer (for generating unified discriminative features); "YOLOv7 visual detection library" is a lightweight and efficient object detection framework adapted to the Python environment; "low-altitude signal standard function library" refers to a knowledge base such as feature matching templates and contour estimation rules generated based on historical small target images, used to enhance the recognition ability of low-altitude slow-moving targets.

[0030] Understandably, single-modal detection methods are unstable in multi-source heterogeneous environments, especially in low-light, occluded, or complex background areas. This step, however, achieves effective detection of small, slow-moving, and sparsely modal flying targets in the low-altitude urban area by fusing information from different modalities into a deep detection network and combining it with the rapid recognition capabilities of the YOLOv7 detector.

[0031] Step S40: Based on the preliminary target set Establish the target state vector And based on the target state vector Predicting the i-th target at the next time step using the extended Kalman filter method +1 target state The target state is tracked using an RF drift compensation tracking mechanism. Perform continuous predictive tracking and output the trajectory estimation result of the i-th target; It should be noted that the "target state vector" includes multi-dimensional information such as target position, velocity, signal strength, and track offset angle; the "extended Kalman filter" is used for nonlinear state prediction, and continuously corrects the prediction error based on the Bayesian framework; the "RF drift compensation tracking mechanism" refers to modeling the evolution trend of the radio frequency signal using frequency offset drift and phase perturbation between consecutive frames, and then compensating in reverse to improve the continuity and accuracy of the trajectory.

[0032] Understandably, this step can continuously predict the target state through the frequency domain compensation mechanism of the radio frequency signal during the detection discontinuity or signal attenuation stage, thereby avoiding interruption or drift of the target trajectory and improving the overall tracking continuity and accuracy.

[0033] It should be understood that, unlike traditional Kalman filtering or pure image trajectory fitting, this invention introduces an RF feature drift compensation mechanism, which can maintain trajectory tracking accuracy and improve target re-identification capability even when the target is temporarily occluded, the image is blurred, or there is optical mismatch. For example, in nighttime tracking tests of long-distance flying targets, the average trajectory prediction error decreased and the effective trajectory reconstruction rate increased after adding RF drift compensation.

[0034] Step S50: Based on the trajectory estimation results, a confidence verification and constant false alarm rate fusion mechanism is used to trigger an effective detection alarm event.

[0035] It should be noted that the "confidence verification mechanism" refers to constructing a target confidence factor by using the consistency between the trajectory estimate and the target's historical state, as well as the trend of signal feature changes; the "constant false alarm rate (CFAR)" mechanism is used to dynamically adjust the detection threshold under a set false alarm probability to avoid false alarms caused by background interference; the two are combined to form a dynamic alarm trigger threshold, which outputs a valid alarm event when the confidence factor exceeds the threshold.

[0036] Understandably, fixed thresholds or static judgment mechanisms often fail to meet the low signal-to-noise ratio target detection requirements in the dynamic urban environment, leading to frequent false alarms. This invention, by integrating confidence dynamic verification with a CFAR threshold adaptive mechanism, significantly reduces the false alarm rate, effectively balancing the trade-off between high detection sensitivity and low false alarm rate, ensuring accurate identification of real threat targets in complex scenarios.

[0037] Example 2: Furthermore, the present invention provides a low-altitude economic airspace passive safety early warning system, employing an early warning method for low-altitude economic airspace passive safety as described in the above embodiments, which can solve the technical problem of low-altitude economic airspace passive safety early warning. Compared with the prior art, the beneficial effects of the low-altitude economic airspace passive safety early warning system provided by the present invention are the same as the beneficial effects of the low-altitude economic airspace passive safety early warning method provided in the above embodiments, and other technical features of the low-altitude economic airspace passive safety early warning system are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0038] Example 3: This invention provides an early warning device for passive safety in low-altitude economic airspace. Please refer to... Figure 2A low-altitude economic airspace passive safety warning device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to execute the low-altitude economic airspace passive safety warning method described in Embodiment 1 above. The low-altitude economic airspace passive safety warning device in this embodiment may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), vehicle terminals (e.g., vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. This low-altitude economic airspace passive safety warning device is merely an example and should not impose any limitations on the functionality and scope of use of this embodiment. The low-altitude economic airspace passive safety warning device may include a processing device 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory 1002 or a program loaded from a storage device 1003 into a random access memory 1004. Random access memory 1004 also stores various programs and data required for the operation of a low-altitude economic airspace passive safety warning device. Processing unit 1001, read-only memory 1002, and random access memory 1004 are interconnected via bus 1005. I / O interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows a low-altitude economic airspace passive safety warning device to communicate wirelessly or wiredly with other devices to exchange data. Although a low-altitude economic airspace passive safety warning device with various systems is shown in the figure, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems may be implemented or possessed alternatively.

[0039] Example 4: This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the aforementioned early warning method for passive safety in low-altitude economic airspace. The computer program product provided by this invention can solve the technical problem of early warning for passive safety in low-altitude economic airspace. Compared with the prior art, the beneficial effects of the computer program product provided by this invention are the same as those of the early warning method for passive safety in low-altitude economic airspace provided in the above embodiments, and will not be repeated here.

[0040] In particular, according to the embodiments disclosed in this invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this invention.

[0041] It should be understood that the various parts disclosed in this invention can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0042] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for early warning of passive safety in low-altitude economic airspace, characterized in that, The methods include: Step S10: By deploying a sensor array in the core urban area, real-time data is acquired based on the sensor array at any given time. The observation data stream is processed; clock synchronization is performed on the observation data stream using a combination of GNSS terminal timing and NTP synchronization protocol, outputting a unified timestamp; based on the unified timestamp, cross-correlation analysis is used to perform time deviation compensation processing on the observation data stream, outputting multimodal frame data. ; Step S20: Obtain GIS building model data and sensor view parameters, and construct an occlusion map based on the GIS building model data and sensor view parameters. Based on occlusion map and multimodal frame data A spatial confidence distribution map is established using a priori perception capability estimation mechanism. And based on the spatial confidence distribution map Determine the priority modal frame set ; Step S30: Set the priority modal frames Input to a pre-defined multimodal target detection network Multimodal target detection network Output fusion discriminative features Based on fusion discriminative features The YOLOv7 vision detection library in Python and the pre-defined low-altitude signal standard function library are used to perform a multi-channel small target detection task, and a preliminary target set is output. ; Step S40: Based on the preliminary target set Establish the target state vector And based on the target state vector Predicting the i-th target at the next time step using the extended Kalman filter method +1 target state The target state is tracked using an RF drift compensation tracking mechanism. Perform continuous predictive tracking and output the trajectory estimation result of the i-th target; Step S50: Based on the trajectory estimation results, a confidence verification and constant false alarm rate fusion mechanism is used to trigger an effective detection alarm event.

2. The early warning method for passive safety in low-altitude economic airspace as described in claim 1, characterized in that, In step S10, the sensor array includes a radar sensor, an infrared thermal imaging sensor, a visible light sensor, and an RF receiver.

3. The early warning method for passive safety in low-altitude economic airspace as described in claim 1, characterized in that, In step S10, the observation data stream includes millimeter-wave radar echo signals. Thermal infrared image frame sequence RGB video image frames Channel status information of RF spectrum .

4. The early warning method for passive safety in low-altitude economic airspace as described in claim 1, characterized in that, In step S20, GIS building model data and sensor view parameters are obtained, and an occlusion map is constructed based on the GIS building model data and sensor view parameters. Based on occlusion map and multimodal frame data A spatial confidence distribution map is established using a priori perception capability estimation mechanism. And based on the spatial confidence distribution map Determine the priority modal frame set The steps specifically include: Step S201: Obtain GIS building model data and sensor view parameters, and construct an occlusion map based on the GIS building model data and sensor view parameters using the view volume projection method. ; Occlusion image Used to characterize whether each point in space is within the visible area; Step S202: Introduce a perception intensity mapping function, which is used to estimate the perception probability of different modal data at different points in three-dimensional space; based on the occlusion map... and multimodal frame data By combining the perception intensity mapping function, the comprehensive observability score at each three-dimensional spatial point is calculated; Step S203: When the overall observability score is greater than the preset perception score threshold, the three-dimensional spatial point is determined to be a recognizable region, and the priority modality frame set is finally output. .

5. The early warning method for passive safety in low-altitude economic airspace as described in claim 1, characterized in that, In step S30, the multimodal target detection network Includes a multimodal input normalization layer for receiving a priority modal frame set. Furthermore, by performing size resampling, dynamic range normalization, and channel alignment operations, the priority modal frame set is... Unify the tensor representation to a unified form, forming a fused input tensor; The modality feature extraction layer is used to encode the features of different modality tensors in the fused input tensor to obtain modality feature maps; the modality attention fusion layer is used to perform weighted fusion based on the modality feature maps using an attention mechanism to generate fused feature maps. The multi-scale detection decoding layer is used to perform spatial scale decoding based on the fused feature map and output fused discriminative features. .

6. The early warning method for passive safety in low-altitude economic airspace as described in claim 1, characterized in that, In step S40, an RF drift compensation tracking mechanism is used to track the target state. The steps for performing continuous predictive tracking and outputting the trajectory estimation result of the i-th target specifically include: acquiring the target state from the RF receiver. The corresponding Channel State Change Feature (CSI) is used to extract the center drift rate of the CSI within a preset sliding window, and the dynamic fusion coefficient is set based on the center drift rate. , dynamic fusion coefficient Applied to target state Perform a weighted combination and output the trajectory estimation result of the i-th target.

7. The early warning method for passive safety in low-altitude economic airspace as described in claim 1, characterized in that, Step S50, which involves triggering an effective detection alarm event based on the trajectory estimation result using a confidence verification and constant false alarm rate (CFAR) fusion mechanism, specifically includes: introducing the CFAR framework and setting a dynamic confidence threshold based on the CFAR framework. Obtain the spatial confidence score corresponding to the trajectory estimation result. When the spatial confidence score Greater than or equal to the dynamic confidence threshold When a valid alarm event is detected, it is determined that there is a valid alarm event. Valid alarm events include illegal low-altitude flight alarm events, static camouflage target drift and sudden change alarm events, target reappearance after continuous obscuration alarm events, and electromagnetic disturbance source alarm events; and the valid alarm events are reported to the regional monitoring platform.

8. A low-altitude economic airspace passive safety early warning system, applied to the low-altitude economic airspace passive safety early warning method according to any one of claims 1 to 7, characterized in that, The low-altitude economic airspace passive safety early warning system includes: The spatiotemporal synchronization and data preprocessing module is used to acquire real-time data based on sensor arrays deployed in the city's core area. The observation data stream is processed; clock synchronization is performed on the observation data stream using a combination of GNSS terminal timing and NTP synchronization protocol, outputting a unified timestamp; based on the unified timestamp, cross-correlation analysis is used to perform time deviation compensation processing on the observation data stream, outputting multimodal frame data. ; The view modeling and perception confidence distribution construction module is used to acquire GIS building model data and sensor view parameters, and to construct an occlusion map based on the GIS building model data and sensor view parameters. Based on occlusion map and multimodal frame data A spatial confidence distribution map is established using a priori perception capability estimation mechanism. And based on the spatial confidence distribution map Determine the priority modal frame set ; The multimodal target detection and feature fusion module is used to combine the priority modal frame set. Input to a pre-defined multimodal target detection network Multimodal target detection network Output fusion discriminative features Based on fusion discriminative features The YOLOv7 vision detection library in Python and the pre-defined low-altitude signal standard function library are used to perform a multi-channel small target detection task, and a preliminary target set is output. ; The target tracking and RF drift compensation prediction module is used to predict targets based on a preliminary target set. Establish the target state vector And based on the target state vector Predicting the i-th target at the next time step using the extended Kalman filter method +1 target state The target state is tracked using an RF drift compensation tracking mechanism. Perform continuous predictive tracking and output the trajectory estimation result of the i-th target; The trajectory verification and alarm triggering module is used to trigger effective detection alarm events based on the trajectory estimation results using a confidence verification and constant false alarm rate fusion mechanism.

9. A passive safety early warning device for low-altitude economic airspace, characterized in that, The low-altitude economic airspace passive safety early warning device includes: a memory, a processor, and a low-altitude economic airspace passive safety early warning program stored in the memory and executable on the processor. When the low-altitude economic airspace passive safety early warning program is executed by the processor, it implements a low-altitude economic airspace passive safety early warning method according to any one of claims 1 to 7.

10. A computer program product, characterized in that, The computer program product includes a low-altitude economic airspace passive safety early warning program, which, when executed by a processor, implements a low-altitude economic airspace passive safety early warning method according to any one of claims 1 to 7.

Images