Multi-source fusion perception vehicle-road cooperation emergency warning method and system

Abstract

The present application belongs to the technical field of intelligent traffic and low-altitude economy integration, in particular to a multi-source fusion perception vehicle-road cooperation emergency warning method and system, the system comprising a perception layer, a communication layer and a control layer; through multi-source data fusion, the situation of the monitoring area is accurately judged, and then a warning signal and a cleaning signal are sent out. Through the fusion of the base station perception data of the 5.5G base station, the radar data of the 92GHz millimeter wave radar and the visual data of the camera, the monitoring range can be expanded to 3.5 kilometers, and when an anomaly is monitored, emergency warning can be carried out in time.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation and low-altitude economy integration technology, and in particular to a multi-source fusion perception vehicle-road cooperative emergency alarm method and system. Background Technology

[0002] When roads collapse or bridges deform, failure to address the issue promptly can pose potential safety risks to passing vehicles and pedestrians. Monitoring road or bridge conditions typically relies on single sensors such as cameras or traditional millimeter-wave radar. However, traditional millimeter-wave radar (e.g., 24GHz) has a short detection range (<100 meters), insufficient for large-scale monitoring such as bridge slopes. 5G base stations have limited sensing capabilities (300-meter coverage), and time-sharing scheduling for communication / sensing results in high latency. Cameras are susceptible to weather interference, leading to insufficient accuracy in deformation monitoring.

[0003] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of the present invention, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to solve the technical problems existing in the background art. To this end, a multi-source fusion perception vehicle-road cooperative emergency alarm method and system are provided to realize closed-loop management of vehicle tracking, infrastructure deformation monitoring and emergency communication.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A vehicle-road cooperative emergency alarm method based on multi-source data fusion perception includes the following steps:

[0007] Step S1: The 5.5G base station switches to the 28GHz frequency band to transmit LFM pulse waves and scans the monitoring area, which includes road areas and bridge areas, to obtain base station sensing data;

[0008] Step S2: If structural deformation occurs in the monitored area, activate the 92GHz millimeter-wave radar and camera;

[0009] Step S3: Obtain radar data of the corresponding monitoring area through 92GHz millimeter-wave radar, and obtain visual data of the corresponding monitoring area through camera;

[0010] Step S4: Perform data fusion of base station sensing data and radar data, and combine it with visual data to obtain the monitoring results for the corresponding monitoring area;

[0011] Step S5: Based on the monitoring results, the 5.5G base station sends a signal to the monitoring center, the monitoring center sends a cleanup command to the cleanup workstation, and the staff at the cleanup workstation go to the corresponding monitoring area to clean up based on the information in the cleanup command.

[0012] Meanwhile, the 5.5G base station uses its communication function to track the location information of users who have registered and logged into the system in real time, and determines the distance between their location and the corresponding monitoring area;

[0013] Step S6: If the distance is less than the distance threshold, an alarm signal is sent to the user and displayed on the interface or the user's mobile phone is triggered to make a call; if the distance is greater than the distance threshold, the monitoring results of the corresponding monitoring area are pushed to the user's message bar.

[0014] The following is a further defined technical solution for the method in this invention: the frequency division sensing range of the 5.5G base station is 2 kilometers, and the detection range of the 92GHz millimeter-wave radar is 1.4 kilometers. The data fusion process of the base station sensing data and the radar data includes the following steps:

[0015] Step 1: Data Synchronization and Spatiotemporal Registration: The GPS synchronization signal from the 5.5G base station is used to provide a unified timestamp for the 92GHz radar and camera, with time error controlled within μs. The radar data and the base station sensing data are aligned through a sliding time window mechanism with a window length of 10ms to ensure accurate matching of the dynamic target's trajectory. Spatial Registration: A multi-source calibration algorithm is used to convert the polar coordinate system data of the radar data and the Cartesian coordinate system data of the base station sensing data to the same spatial reference system. Joint calibration is performed using a calibration board, and the spatial registration error is ≤0.1 meters.

[0016] Step 2, Feature extraction of base station sensing data: LFM-OFDM hybrid waveform is used, with the 3.5GHz band responsible for near-end communication and the 28GHz band for long-range sensing, with a pulse width of 100ns and a bandwidth of 100MHz, achieving 2km coverage; target speed and azimuth are extracted by Doppler frequency shift and angle of arrival of the echo signal.

[0017] Feature extraction from radar data: Point cloud generation, high-density point cloud is generated by processing echo signals through FFT, and target contours and micro-Doppler features are extracted;

[0018] Step 3, Data-level Fusion: The original radar point cloud and the base station sensing signal are spatiotemporally superimposed, an adaptive Kalman filter is used to eliminate multipath interference from moving targets, the state equation is fused to obtain the target position and velocity, and the covariance matrix is ​​dynamically updated.

[0019] Step 4, Decision-level fusion: The target attributes are synthesized using DS evidence theory, and the basic probability allocation is used to finally output a high fusion confidence score.

[0020] The following is a further defined technical solution of the method in this invention: based on the monitoring results, the road surface collapse mode of the road area is determined, including: a 92GHz millimeter-wave radar monitors the micro-deformation of the road surface in the road area through interferometry technology, identifies underground cavities or voids, calculates the actual deformation amount by combining phase unwrapping algorithm, and dynamically updates the risk coefficient model; a multispectral high-definition camera identifies road surface collapse pits in real time based on the YOLOv5-CS model, classifies them as "ordinary collapse" or "dangerous collapse", calculates the actual area of ​​the collapse by combining camera parameters, and triggers a threshold warning; a 5.5G base station integrates radar data and camera visual data, eliminates environmental interference through multi-source fusion algorithm, and outputs the collapse risk level.

[0021] The following is a further technical solution for the method of this invention. The road collapse mode has three levels, including primary warning, intermediate warning and emergency mode. The primary warning includes radar detection of a settlement depth of less than 1 mm or camera recognition of a collapse area of ​​less than 5 m², triggering routine inspection and re-measurement. The intermediate warning includes underground cavities with a depth greater than 0.5 m or a collapse area of ​​5-10 m², initiating grouting reinforcement and temporary traffic control. The emergency mode includes a collapse area greater than 10 m² or pipeline leakage, immediately closing the road section and coordinating with the emergency command center to initiate grouting and concrete backfilling emergency measures.

[0022] The following is a further technical solution for the method in this invention: based on the monitoring results, the bridge deformation mode in the bridge area is determined, including: monitoring the deflection and vibration frequency of the main beam with 92GHz millimeter-wave radar, identifying whether the fundamental frequency shift of the structural mode is greater than 5% through spectrum analysis, and uploading the dynamic deflection data to the edge computing node in real time for Kalman filtering noise reduction.

[0023] The following is a further technical solution for the method in this invention: the bridge deformation mode has three levels, including primary warning, intermediate warning and emergency mode. The primary warning includes elastic deformation reaching 0.1‰ of the span or vibration frequency fluctuation exceeding 5%, triggering routine inspection; the intermediate warning includes plastic deformation exceeding 1‰ of the span or abnormal damping ratio, initiating structural reinforcement; the emergency mode includes deformation greater than or equal to 2‰ of the span or sudden change in cable force, immediately closing the bridge and activating the backup passage.

[0024] The following is a further technical solution for the method of the present invention: In the emergency mode of road surface collapse mode in the road area or in the emergency mode of bridge deformation mode in the bridge area, the 5.5G base station enters the fish scale networking mode. The fish scale networking mode uses a hybrid wave combining continuous wave OFDM and pulse wave LFM to track the user's location information in real time.

[0025] A vehicle-road cooperative emergency alarm system based on multi-source data fusion and perception, used to implement the above method, includes:

[0026] The perception layer includes a 92GHz millimeter-wave radar array and cameras. The 92GHz millimeter-wave radar array supports a 1.4km scanning coverage and acquires radar data for the corresponding monitoring area. The cameras enable license plate or user biometric recognition and acquire visual data for the corresponding monitoring area.

[0027] The communication layer includes 5.5G base stations, which monitor all monitoring areas and acquire base station sensing data from all monitoring areas.

[0028] The control layer performs multi-source data fusion processing.

[0029] Compared with the prior art, the present invention has the following technical effects:

[0030] This invention integrates base station sensing data from a 5.5G base station, radar data from a 92GHz millimeter-wave radar, and visual data from a camera, which can extend the monitoring range to 3.5 kilometers. When an anomaly is detected, an emergency alarm can be issued in a timely manner.

[0031] The present invention will be further described below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in 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.

[0033] Figure 1 This is a schematic diagram of the system flow of the present invention. Detailed Implementation

[0034] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0035] like Figure 1 As shown, a vehicle-road cooperative emergency alarm system based on multi-source data fusion perception is provided, which mainly consists of a perception layer, a communication layer, and a control layer.

[0036] The perception layer comprises a 92GHz millimeter-wave radar array and cameras. The 92GHz millimeter-wave radar array (for guardrail / tunnel-specific) supports 1.4km scanning coverage, acquiring radar data for the corresponding monitoring area. High-definition cameras (including infrared modules) enable license plate or user biometric recognition, acquiring visual data for the corresponding monitoring area. The communication layer includes a 5.5G base station with a 60° vertical angle and a coverage height of 300 meters, monitoring all monitoring areas and acquiring base station perception data for all areas. The control layer (application server and monitoring center) performs multi-source data fusion processing. The 92GHz millimeter-wave radar array achieves ultra-long-range detection of 1.4km and 0.1m-level resolution, accurately capturing micro-deformations of bridges. The 5.5G base station's frequency division detection uses LFM-OFDM hybrid waveforms to achieve parallel communication (3.5GHz) and perception (28GHz), extending the detection range to 2km. Multimodal fusion analysis of radar point cloud, visual recognition, and base station perception data reduces the false alarm rate by 60%.

[0037] A vehicle-road cooperative emergency alarm method based on multi-source data fusion perception includes the following steps:

[0038] Step S1: The 5.5G base station switches to the 28GHz band to transmit LFM pulse waves and scans the road or bridge deformation in the key monitoring area (sensitivity ±0.5mm). The monitoring area includes the road area and the bridge area, and the base station sensing data is obtained.

[0039] Step S2: If structural deformation occurs in the monitored area, i.e., after the 5.5G base station detects a potential hazard, the 92GHz millimeter-wave radar and camera are activated to confirm the risk.

[0040] Step S3: Obtain radar data for the corresponding monitoring area through 92GHz millimeter-wave radar, obtain visual data for the corresponding monitoring area through camera, and transmit data back to the monitoring center in real time through 5.5G base station via 3.5GHz frequency band.

[0041] Step S4: Perform data fusion of base station sensing data and radar data, and combine it with visual data to obtain the monitoring results for the corresponding monitoring area.

[0042] Step S5: Based on the monitoring results (road collapse or bridge deformation), the 5.5G base station sends a signal to the monitoring center, which then issues a cleanup command to the cleanup workstation. The staff at the cleanup workstation then proceed to the corresponding monitoring area to carry out the cleanup based on the information in the cleanup command.

[0043] Meanwhile, the 5.5G base station uses its communication function to track the location information of users who have registered and logged into the system in real time, and determines the distance between their location and the corresponding monitoring area.

[0044] Step S6: If the distance is less than the distance threshold, an alarm signal is sent to the user and displayed on the interface or the user's mobile phone is triggered to make a call (delay <10ms); if the distance is greater than the distance threshold, the monitoring results of the corresponding monitoring area are pushed to the user's message bar.

[0045] The data fusion process of base station sensing data and radar data includes the following steps:

[0046] Step 1: Data Synchronization and Spatiotemporal Registration: The GPS synchronization signal (PTP protocol) of the 5.5G base station is used to provide a unified timestamp for the 92GHz radar and camera, with the time error controlled within the μs level; the radar data (pulse sampling rate 1GHz) and the base station sensing data (28GHz band, sampling rate 500MHz) are aligned through a sliding time window mechanism with a window length of 10ms to ensure accurate matching of the dynamic target trajectory; Spatial Registration: A multi-source calibration algorithm is used to convert the polar coordinate system data (range-azimuth) of the radar data and the Cartesian coordinate system data (XYZ) of the base station sensing data to the same spatial reference system, and joint calibration is performed through a calibration board, with a spatial registration error ≤0.1m.

[0047] Step 2, Feature extraction of base station sensing data: LFM-OFDM hybrid waveform is used, with the 3.5GHz band (continuous wave) responsible for near-end communication and the 28GHz band (pulse wave) for long-range sensing. The pulse width is 100ns and the bandwidth is 100MHz, achieving 2km coverage. The target speed and azimuth are extracted by the Doppler frequency shift and angle of arrival of the echo signal.

[0048] Feature extraction from radar data: Point cloud generation, high-density point cloud is generated by processing echo signals through FFT, and target contours and micro-Doppler features are extracted.

[0049] Step 3, Data-level Fusion: The original radar point cloud and the base station sensing signal are spatiotemporally superimposed, and an adaptive Kalman filter is used to eliminate multipath interference from moving targets. The state equation is fused with the target position (X, Y, Z) and velocity (Vx, Vy, Vz), and the covariance matrix is ​​dynamically updated.

[0050] Step 4, Decision-level fusion: Target attributes (such as vehicles / bridges) are synthesized using DS evidence theory, and basic probability allocation is used to finally output a high fusion confidence score.

[0051] Based on monitoring results, the road surface subsidence pattern is determined, including: 92GHz millimeter-wave radar uses interferometry to monitor micro-deformation of the road surface, identify underground cavities or voids, calculate the actual deformation using a phase unwrapping algorithm, and dynamically update the risk coefficient model; multispectral high-definition cameras use the YOLOv5-CS model to identify road subsidence pits in real time, classifying them as "ordinary subsidence" or "dangerous subsidence," and calculate the actual subsidence area based on camera parameters to trigger a threshold warning; 5.5G base stations integrate radar data and camera visual data, eliminate environmental interference through a multi-source fusion algorithm, and output the subsidence risk level. The road surface subsidence pattern has three levels: primary warning, intermediate warning, and emergency mode. The alert levels are categorized into three levels: Primary alert (blue / yellow) includes radar detection of subsidence less than 1 mm or camera identification of collapses less than 5 square meters, triggering routine patrols and re-monitoring; Intermediate alert (orange) includes underground cavities deeper than 0.5 m or collapses covering an area of ​​5-10 square meters, initiating grouting reinforcement and temporary traffic control; Emergency mode (red) includes collapses covering an area greater than 10 square meters or pipeline leaks, immediately closing the road section and coordinating with the emergency command center to initiate grouting and concrete backfilling rescue measures. Long-term optimization includes establishing a collapse hazard database and dynamically adjusting monitoring thresholds.

[0052] Based on monitoring results, the bridge deformation mode in the bridge area is determined, including: monitoring the deflection and vibration frequency of the main girder with 92GHz millimeter-wave radar; identifying whether the fundamental frequency shift of the structural modes exceeds 5% through spectrum analysis; and uploading dynamic deflection data to edge computing nodes in real time for Kalman filtering noise reduction. The bridge deformation mode has three levels: primary warning, intermediate warning, and emergency mode. Primary warning (blue) includes elastic deformation exceeding 0.1‰ of the span or vibration frequency fluctuation exceeding 5%, triggering routine inspections; intermediate warning (orange) includes plastic deformation exceeding 1‰ of the span or abnormal damping ratio, initiating structural reinforcement; and emergency mode (red) includes deformation greater than or equal to 2‰ of the span or sudden cable force changes, immediately closing the bridge and activating alternative passageways. Traffic control: Heavy vehicles are prohibited from passing, and detour information is disseminated in real time via 5.5G base stations.

[0053] In emergency modes such as road surface collapse or bridge deformation, 5.5G base stations enter a fish-scale networking mode. This mode uses a hybrid wave combining continuous wave OFDM and pulse wave LFM to track user location information in real time. Unlike traditional cellular base stations that use three 120-degree sectors, all base station sectors in a fish-scale network face the same direction, arranged in a layered pattern resembling fish scales. Based on "integrated sensing and communication" technology, the fish-scale network enables base stations to not only possess traditional communication functions but also sense the location and deformation of detected targets, much like radar. Traditional 5G base stations, operating in a self-transmitting and self-receiving mode, have limited transmission power due to isolation constraints, typically achieving only about 300 meters of coverage. LFM waveforms, through time-division multiplexing of transmission and reception, overcome this limitation, extending single-station coverage from hundreds of meters to kilometers. Fish-scale network base stations are specially designed, increasing their vertical angle from 24° (traditional base stations) to over 60°, and their coverage height from 100 meters to 300 meters.

[0054] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any person skilled in the art can make many possible variations and modifications to the technical solution of the present invention, or modify it into equivalent embodiments, without departing from the scope of the present invention's technical solution. Therefore, all equivalent changes made based on the shape, structure, and principle of the present invention without departing from the scope of the present invention's technical solution should be covered within the protection scope of the present invention.

Claims

1. A vehicle-road cooperative emergency alarm method based on multi-source data fusion and perception, characterized in that, Includes the following steps: Step S1: The 5.5G base station switches to the 28GHz frequency band to transmit LFM pulse waves and scans the monitoring area, which includes road areas and bridge areas, to obtain base station sensing data; Step S2: If structural deformation occurs in the monitored area, activate the 92GHz millimeter-wave radar and camera; Step S3: Obtain radar data of the corresponding monitoring area through 92GHz millimeter-wave radar, and obtain visual data of the corresponding monitoring area through camera; Step S4: Perform data fusion of base station sensing data and radar data, and combine it with visual data to obtain the monitoring results for the corresponding monitoring area; Step S5: Based on the monitoring results, the 5.5G base station sends a signal to the monitoring center, the monitoring center sends a cleanup command to the cleanup workstation, and the staff at the cleanup workstation go to the corresponding monitoring area to clean up based on the information in the cleanup command. Meanwhile, the 5.5G base station uses its communication function to track the location information of users who have registered and logged into the system in real time, and determines the distance between their location and the corresponding monitoring area; Step S6: If the distance is less than the distance threshold, an alarm signal is sent to the user and displayed on the interface or the user's mobile phone is triggered to make a call; if the distance is greater than the distance threshold, the monitoring results of the corresponding monitoring area are pushed to the user's message bar. The frequency division sensing range of a 5.5G base station is 2 kilometers, while the detection range of a 92GHz millimeter-wave radar is 1.4 kilometers. The data fusion process between base station sensing data and radar data includes the following steps: Step 1: Data Synchronization and Spatiotemporal Registration: The GPS synchronization signal from the 5.5G base station is used to provide a unified timestamp for the 92GHz radar and camera, with time error controlled within μs. The radar data and the base station sensing data are aligned through a sliding time window mechanism with a window length of 10ms to ensure accurate matching of the dynamic target's trajectory. Spatial Registration: A multi-source calibration algorithm is used to convert the polar coordinate system data of the radar data and the Cartesian coordinate system data of the base station sensing data to the same spatial reference system. Joint calibration is performed using a calibration board, and the spatial registration error is ≤0.1 meters. Step 2, Feature extraction of base station sensing data: LFM-OFDM hybrid waveform is used, with the 3.5GHz band responsible for near-end communication and the 28GHz band for long-range sensing, with a pulse width of 100ns and a bandwidth of 100MHz, achieving 2km coverage; target speed and azimuth are extracted by Doppler frequency shift and angle of arrival of the echo signal. Feature extraction from radar data: Point cloud generation, high-density point cloud is generated by processing echo signals through FFT, and target contours and micro-Doppler features are extracted; Step 3, Data-level Fusion: The original radar point cloud and the base station sensing signal are spatiotemporally superimposed, an adaptive Kalman filter is used to eliminate multipath interference from moving targets, the state equation is fused to obtain the target position and velocity, and the covariance matrix is ​​dynamically updated. Step 4, Decision-level fusion: The target attributes are synthesized using DS evidence theory, and the basic probability allocation is used to finally output a high fusion confidence score.

2. The vehicle-road cooperative emergency alarm method based on multi-source data fusion perception as described in claim 1, characterized in that, Based on the monitoring results, the road surface collapse pattern in the road area is determined, including: 92GHz millimeter-wave radar monitors the micro-deformation of the road surface in the road area through interferometry technology, identifies underground cavities or voids, calculates the actual deformation amount by combining phase unwrapping algorithm, and dynamically updates the risk coefficient model; multispectral high-definition cameras identify road surface collapse pits in real time based on the YOLOv5-CS model, classifying them as "ordinary collapse" or "dangerous collapse", calculate the actual area of ​​the collapse by combining camera parameters, and trigger threshold warnings; 5.5G base stations integrate radar data and camera visual data, eliminate environmental interference through multi-source fusion algorithms, and output the collapse risk level.

3. The vehicle-road cooperative emergency alarm method based on multi-source data fusion perception as described in claim 2, characterized in that, The road collapse mode has three levels, including primary warning, intermediate warning and emergency mode. Primary warning includes radar detection of subsidence depth of less than 1 mm or camera recognition of collapse area of ​​less than 5 square meters, triggering routine inspection and re-measurement. Intermediate warning includes underground cavity depth greater than 0.5 m or collapse area of ​​5-10 square meters, initiating grouting reinforcement and temporary traffic control. Emergency mode includes collapse area greater than 10 square meters or pipeline leakage, immediately closing the road section and coordinating with the emergency command center to initiate grouting and concrete backfilling emergency measures.

4. The vehicle-road cooperative emergency alarm method based on multi-source data fusion perception as described in claim 1, characterized in that, Based on the monitoring results, the deformation mode of the bridge in the bridge area is determined, including: monitoring the deflection and vibration frequency of the main beam by 92GHz millimeter-wave radar, identifying whether the fundamental frequency shift of the structural mode is greater than 5% through spectrum analysis, and uploading dynamic deflection data to the edge computing node in real time for Kalman filtering noise reduction.

5. The vehicle-road cooperative emergency alarm method based on multi-source data fusion perception as described in claim 4, characterized in that, The bridge deformation mode has three levels, including primary warning, intermediate warning and emergency mode. Primary warning includes elastic deformation of 0.1‰ of the span or vibration frequency fluctuation of more than 5%, triggering routine inspection; intermediate warning includes plastic deformation of more than 1‰ of the span or abnormal damping ratio, initiating structural reinforcement; emergency mode includes deformation of more than or equal to 2‰ of the span or sudden change in cable force, immediately closing the bridge and activating the backup passage.

6. The vehicle-road cooperative emergency alarm method based on multi-source data fusion perception as described in claim 1, characterized in that, In emergency mode of road surface collapse or bridge deformation in bridge area, 5.5G base stations enter fish-scale networking mode. Fish-scale networking mode uses a hybrid wave combining continuous wave OFDM and pulse wave LFM to track user location information in real time.

7. A vehicle-road cooperative emergency alarm system based on multi-source data fusion sensing, used to implement the method described in any one of claims 1-6, characterized in that, include: The perception layer includes a 92GHz millimeter-wave radar array and cameras. The 92GHz millimeter-wave radar array supports a 1.4km scanning coverage and acquires radar data for the corresponding monitoring area. The cameras enable license plate or user biometric recognition and acquire visual data for the corresponding monitoring area. The communication layer includes 5.5G base stations, which monitor all monitoring areas and acquire base station sensing data from all monitoring areas. The control layer performs multi-source data fusion processing.

Images