A car-road cooperation RSU terminal based on real-time information of a field road

Abstract

The application discloses the vehicle-road cooperation RSU terminal technical field, specifically a kind of vehicle-road cooperation RSU terminal based on field road real-time information, first integrated multi-class sensor synchronous acquisition road full-dimension information, and through environmental self-adaptive calibration adaptation bad weather and complex scene, and by target feature purification eliminates invalid data, provide accurate, comprehensive raw data support for subsequent processing;Subsequently, with light weight federated learning technology, privacy protection and distributed computing needs are considered, and through real-time data cleaning, the multi-source data synchronization problem is solved, and a dynamic priority scheduling mechanism is matched, emergency data is processed preferentially;Based on this, the application cooperates and adaptively calibrates through multi-source heterogeneous sensor, enriches the perception dimension, completely breaks the limitation of single perception mode, can calmly cope with the interference brought by bad weather and complex road conditions, while ensuring the accuracy of perception data through target feature purification.

Description

Technical Field

[0001] This invention relates to the field of vehicle-road cooperative RSU terminal technology, specifically a vehicle-road cooperative RSU terminal based on real-time on-site road information. Background Technology

[0002] With the rapid development of the intelligent connected vehicle industry, vehicle-to-everything (V2X) has become a core support for improving traffic safety and efficiency. Roadside units (RSUs), as key devices for roadside perception, data processing, and information exchange, are widely deployed in scenarios such as intersections, tunnel entrances, and dangerous road sections. Currently, RSU terminals rely on technologies such as 5G-V2X, millimeter-wave radar, and high-definition cameras to achieve real-time data exchange between vehicles and the road environment, playing an important role in scenarios such as blind spot warning, coordinated lane changing, and congestion mitigation. However, with the improvement of autonomous driving levels, higher requirements are placed on the real-time performance, anti-interference capabilities, compatibility, and safety of RSU terminals. The existing equipment's adaptability to complex road conditions and extreme environments is insufficient, restricting the large-scale deployment of vehicle-to-everything technology. The specific problems are as follows:

[0003] 1. High data processing latency and insufficient privacy protection.

[0004] Most RSUs (Roadside Units) need to upload collected data to the cloud for processing, resulting in end-to-end latency far exceeding safety thresholds, and posing a risk of privacy breaches during data transmission. For example, in the event of a sudden traffic accident on a highway, traditional RSUs take 2 seconds to upload video data to the cloud for processing, missing the optimal warning window; on a European highway, hackers sniffed unencrypted RSU data and tampered with warning information for construction zones, causing multiple vehicles to enter closed areas.

[0005] 2. Weak security protection, vulnerable to tampering and attacks.

[0006] RSUs face multiple attack threats at the communication, hardware, and data layers. For example, in a pilot project in a city in 2025, hackers tampered with roadside millimeter-wave radar data, causing autonomous vehicles to misjudge the presence of virtual obstacles ahead, resulting in a nearly 40-minute traffic jam on the main road; another attacker replaced the RSU's EEPROM chip, causing the device to continuously send incorrect high-precision map coordinates, causing multiple vehicles to deviate from their lanes. Summary of the Invention

[0007] To address the aforementioned technical problems of high data processing latency and weak security protection, this invention provides the following technical solution:

[0008] A vehicle-road cooperative RSU terminal based on real-time on-site road information, comprising:

[0009] The multi-source fusion perception module is used to integrate multiple types of sensors to simultaneously collect road information in all dimensions. It adapts to severe weather and complex scenarios through environmental adaptive calibration, and removes invalid data through target feature purification, providing accurate and comprehensive raw data support for subsequent processing.

[0010] The intelligent edge processing module is used to balance privacy protection and distributed computing needs with lightweight federated learning technology. It solves the problem of multi-source data synchronization through real-time data cleaning and is equipped with a dynamic priority scheduling mechanism to prioritize the processing of emergency event data, greatly reducing processing latency and ensuring the efficiency and consistency of data processing.

[0011] The traffic participant intent prediction module is used to focus on three core traffic participants: pedestrians, vehicles, and non-motorized vehicles, to mine their key information, predict risky behaviors, and generate advance warning basis, so that decision-making can be upgraded from passive response to proactive prediction.

[0012] The end-to-end security protection module is used to build a multi-layered protection system from hardware to data. It relies on a dedicated chip to prevent hardware tampering, and uses dedicated line encryption to ensure transmission security. It also monitors abnormal behavior in real time to resist attacks, and is equipped with a security degradation mechanism to maintain core services in extreme situations, thus comprehensively strengthening the terminal security defense.

[0013] The dual-mode collaborative communication module is used to build a cross-protocol adaptation engine to be compatible with mainstream communication standards. It also integrates a dual-mode communication module to deal with signal differences in different scenarios, and is equipped with low-latency transmission technology to achieve stable device docking and efficient information flow across regions and scenarios.

[0014] The dynamic decision-making and scheduling module is used to match real-time road conditions based on preset scenario models, generate precise instructions, and link vehicle terminals, traffic control platforms and surrounding RSUs to achieve three-way collaborative response of vehicles, roads and platforms, quickly handle various traffic scenarios, and improve traffic efficiency and safety.

[0015] The green energy-saving adaptive module is used to dynamically adjust the equipment operating parameters according to traffic density and communication load, and adapt to clean energy to achieve flexible power supply switching. It also monitors the energy consumption data of each module and generates analysis reports to support the maintenance module, taking into account both operating efficiency and energy saving economy.

[0016] The full lifecycle maintenance module is used to monitor the hardware operating status and parameters of each module in real time, and predict the risk of equipment aging and failure through big data analysis, and issue early warnings. It also supports remote firmware upgrades, reducing manual troubleshooting costs, ensuring the continuous and stable operation of the terminal and achieving dynamic performance optimization.

[0017] As a preferred embodiment of the vehicle-road cooperative RSU terminal based on real-time on-site road information described in this invention, the multi-source fusion sensing module includes:

[0018] The multi-source heterogeneous data acquisition unit is used to integrate high-definition cameras, millimeter-wave radar, lidar, meteorological sensors and geomagnetic detectors to simultaneously collect road data in all dimensions.

[0019] An environmental adaptive calibration unit is used to monitor weather and lighting conditions in real time to automatically switch the sensor's operating mode in adverse weather conditions.

[0020] The target feature purification unit is used to initially filter data through edge computing to eliminate false targets caused by changes in light and shadow from the camera, while retaining the accurate feature data of key targets.

[0021] As a preferred embodiment of the vehicle-road cooperative RSU terminal based on real-time road information described in this invention, the intelligent edge processing module includes:

[0022] The Federated Learning Lightweight Unit is used to employ the improved FTD3 algorithm, transmitting only neural network parameters instead of the original data, and enabling distributed computing across RSU clusters.

[0023] The real-time data cleaning unit is used to solve the problem of multi-source data synchronization through outlier removal and timestamp alignment algorithms, keeping the time difference between different sensors within a threshold and ensuring data consistency.

[0024] The dynamic priority scheduling unit is used to classify data processing levels according to the urgency of events, with emergency information being processed first and regular traffic flow data being processed in an orderly manner according to queues.

[0025] As a preferred embodiment of the vehicle-road cooperative RSU terminal based on real-time road information described in this invention, the traffic participant intent prediction module includes:

[0026] The pedestrian behavior prediction unit is used to predict pedestrian behavior based on pedestrian characteristics and the status of traffic lights at intersections, and generate advance warning instructions in advance.

[0027] The vehicle trajectory prediction unit is used to predict vehicle intentions by analyzing vehicle data and combining it with road information.

[0028] The non-motorized vehicle situation prediction unit is used to predict the risky behavior of non-motorized vehicles by integrating geomagnetic detectors and video features.

[0029] As a preferred embodiment of the vehicle-road cooperative RSU terminal based on real-time on-site road information described in this invention, the full-link security protection module includes:

[0030] The hardware anti-tampering unit integrates a physically unclonable function chip and a national cryptographic SM4 algorithm security chip to generate a unique device identity and prevent chip replacement and firmware rewriting attacks.

[0031] The transmission encryption unit is used to encrypt transmission using a 5G-APN leased line and to perform end-to-end encryption on sensed data and control commands to prevent man-in-the-middle attacks and data sniffing.

[0032] The abnormal behavior monitoring unit is used to monitor the frequency and content of data transmission in real time, so as to immediately trigger a security alarm when abnormal data is detected.

[0033] The security degradation unit is used to automatically switch to local offline working mode when encountering a severe attack, so as to provide a minimum early warning service based on pre-stored basic road information and avoid complete system paralysis.

[0034] As a preferred embodiment of the vehicle-road cooperative RSU terminal based on real-time on-site road information described in this invention, the dual-mode cooperative communication module includes:

[0035] Cross-protocol adaptation unit, used for built-in protocol conversion engine, compatible with mainstream protocols, to automatically identify the protocol type of the connected device and complete real-time conversion;

[0036] The dual-mode switching unit is used to integrate 5G-V2X and Beidou short message communication modules to use 5G communication in areas with good signal and automatically switch to Beidou short message mode in areas with no signal.

[0037] Low-latency transmission units are used to shorten data transmission paths by employing edge node proximity access technology.

[0038] As a preferred embodiment of the vehicle-road cooperative RSU terminal based on real-time road information described in this invention, the dynamic decision-making and scheduling module includes:

[0039] The scenario-based decision-making unit is used to preset multiple scenario models to match the optimal decision-making scheme based on real-time traffic conditions and generate decision instructions;

[0040] The multi-object coordination unit is used to simultaneously send coordination commands to the vehicle terminal, traffic control platform and surrounding RSUs to achieve three-way linkage between vehicles, roads and platforms.

[0041] As a preferred embodiment of the vehicle-road cooperative RSU terminal based on real-time on-site road information described in this invention, the green energy-saving adaptive module includes:

[0042] The energy consumption dynamic adjustment unit is used to adjust the sensor operating frequency according to traffic flow density and adjust the transmission power of the dual-mode communication module according to communication load to avoid unnecessary energy consumption.

[0043] The clean energy adapter unit is used to adapt to outdoor clean energy power supply scenarios, monitor the power supply stability in real time, and switch to clean energy power supply first when clean energy is sufficient, and automatically switch to mains power when it is insufficient, thereby reducing energy consumption costs.

[0044] The energy consumption monitoring and feedback unit is used to record the energy consumption data of each module, generate energy consumption analysis reports, and synchronize them to the full life cycle maintenance module to help determine whether the equipment is aging due to abnormal energy consumption, thus forming a collaborative closed loop of energy saving and maintenance.

[0045] As a preferred embodiment of the vehicle-road cooperative RSU terminal based on real-time on-site road information described in this invention, the full lifecycle maintenance module includes:

[0046] The real-time status monitoring unit is used to continuously monitor the working status of the hardware and record the device's operating parameters and energy consumption data;

[0047] The fault prediction and early warning unit is used to analyze the aging trend of equipment based on big data, and send early warning information to the operation and maintenance platform when the performance of components degrades to a threshold, so as to arrange maintenance in advance.

[0048] The remote upgrade unit supports remote OTA upgrades of firmware and algorithms without requiring on-site device disassembly, reducing maintenance costs and continuously optimizing terminal performance.

[0049] Compared with existing technologies:

[0050] This invention effectively addresses the core pain points of existing RSU terminals, possessing comprehensive technical advantages: Through collaborative acquisition by multi-source heterogeneous sensors and adaptive environmental calibration, it enriches the perception dimensions, completely breaking the limitations of a single perception mode. It can readily cope with interference from severe weather and complex road conditions, while target feature purification ensures the accuracy of the perception data. Relying on a cross-protocol adaptation engine and dual-mode communication switching capabilities, it successfully overcomes protocol barriers between different regions and manufacturers, achieving seamless connection and stable communication between devices across scenarios, ensuring the continuity of information interaction for vehicles traveling across regions. Utilizing intelligent edge processing and federated learning lightweight technology, it significantly reduces data processing latency while avoiding… It eliminates the need for external transmission of raw privacy data, balancing data processing efficiency with privacy protection. A multi-layered security system is built through hardware tamper-proofing, end-to-end encrypted transmission, and abnormal behavior monitoring, effectively resisting various attacks such as hardware replacement and data tampering. Meanwhile, a security degradation mechanism ensures the continued operation of core services under extreme conditions. Coupled with a full lifecycle intelligent maintenance mechanism, it enables real-time monitoring of equipment operating status and early fault prediction, eliminating the need for frequent manual troubleshooting. This not only reduces service interruptions caused by equipment downtime but also significantly lowers overall operation and maintenance costs. Furthermore, remote upgrade capabilities ensure continuous optimization of terminal performance, comprehensively supporting the large-scale and stable deployment of vehicle-road cooperative technology. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of the overall framework of the present invention;

[0052] Figure 2 This is a schematic diagram of the multi-source fusion sensing module framework of the present invention;

[0053] Figure 3 This is a schematic diagram of the intelligent edge processing module framework of the present invention;

[0054] Figure 4 This is a schematic diagram of the traffic participant intention prediction module framework of the present invention;

[0055] Figure 5 This is a schematic diagram of the end-to-end security protection module framework of the present invention;

[0056] Figure 6 This is a schematic diagram of the dual-mode collaborative communication module framework of the present invention;

[0057] Figure 7 This is a schematic diagram of the dynamic decision-making and scheduling module framework of the present invention;

[0058] Figure 8 This is a schematic diagram of the green energy-saving adaptive module framework of the present invention;

[0059] Figure 9 This is a schematic diagram of the full lifecycle maintenance module framework of the present invention. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0061] This invention provides a vehicle-road cooperative RSU terminal based on real-time on-site road information. Please refer to [link / reference]. Figure 1 ,include:

[0062] The multi-source fusion perception module is used to integrate multiple types of sensors to simultaneously collect road information in all dimensions. It adapts to severe weather and complex scenarios through environmental adaptive calibration, and removes invalid data through target feature purification, providing accurate and comprehensive raw data support for subsequent processing.

[0063] The intelligent edge processing module is used to balance privacy protection and distributed computing needs with lightweight federated learning technology. It solves the problem of multi-source data synchronization through real-time data cleaning and is equipped with a dynamic priority scheduling mechanism to prioritize the processing of emergency event data, greatly reducing processing latency and ensuring the efficiency and consistency of data processing.

[0064] The traffic participant intent prediction module is used to focus on three core traffic participants: pedestrians, vehicles, and non-motorized vehicles, to mine their key information, predict risky behaviors, and generate advance warning basis, so that decision-making can be upgraded from passive response to proactive prediction.

[0065] The end-to-end security protection module is used to build a multi-layered protection system from hardware to data. It relies on a dedicated chip to prevent hardware tampering, and uses dedicated line encryption to ensure transmission security. It also monitors abnormal behavior in real time to resist attacks, and is equipped with a security degradation mechanism to maintain core services in extreme situations, thus comprehensively strengthening the terminal security defense.

[0066] The dual-mode collaborative communication module is used to build a cross-protocol adaptation engine to be compatible with mainstream communication standards. It also integrates a dual-mode communication module to deal with signal differences in different scenarios, and is equipped with low-latency transmission technology to achieve stable device docking and efficient information flow across regions and scenarios.

[0067] The dynamic decision-making and scheduling module is used to match real-time road conditions based on preset scenario models, generate precise instructions, and link vehicle terminals, traffic control platforms and surrounding RSUs to achieve three-way collaborative response of vehicles, roads and platforms, quickly handle various traffic scenarios, and improve traffic efficiency and safety.

[0068] The green energy-saving adaptive module is used to dynamically adjust the equipment operating parameters according to traffic density and communication load, and adapt to clean energy to achieve flexible power supply switching. It also monitors the energy consumption data of each module and generates analysis reports to support the maintenance module, taking into account both operating efficiency and energy saving economy.

[0069] The full lifecycle maintenance module is used to monitor the hardware operating status and parameters of each module in real time, and predict the risk of equipment aging and failure through big data analysis, and issue early warnings. It also supports remote firmware upgrades, reducing manual troubleshooting costs, ensuring the continuous and stable operation of the terminal and achieving dynamic performance optimization.

[0070] Please see Figure 2 The multi-source fusion sensing module includes:

[0071] The multi-source heterogeneous data acquisition unit is used to integrate high-definition cameras, millimeter-wave radar, lidar, meteorological sensors and geomagnetic detectors to simultaneously collect road data in all dimensions, such as vehicle trajectory, obstacle location, traffic light status, visibility and road surface humidity.

[0072] An environmental adaptive calibration unit is used to monitor weather and lighting conditions in real time, so as to automatically switch the sensor working mode in severe weather conditions such as rain, fog and snow. For example, in heavy rain, infrared cameras and lidar are used for complementary sensing, and radar detection power is enhanced in tunnels.

[0073] The target feature purification unit is used to initially filter data through edge computing to eliminate false targets caused by changes in light and shadow from the camera, while retaining accurate feature data of key targets such as vehicles, pedestrians, and obstacles.

[0074] Please see Figure 3 The intelligent edge processing module includes:

[0075] The Federated Learning Lightweight Unit is used to employ the improved FTD3 algorithm, transmitting only neural network parameters instead of raw data, and enabling distributed computing across RSU clusters, thereby reducing communication overhead and preventing the leakage of privacy data such as vehicle trajectories.

[0076] The real-time data cleaning unit is used to solve the problem of multi-source data synchronization through outlier removal and timestamp alignment algorithms, keeping the time difference between different sensors within a threshold (10 milliseconds) to ensure data consistency.

[0077] The dynamic priority scheduling unit is used to classify data processing levels according to the urgency of events, with emergency information such as traffic accidents and road closures being processed first, while regular traffic flow data is processed in an orderly manner according to queues.

[0078] Please see Figure 4 The traffic participant intention prediction module includes:

[0079] The pedestrian behavior prediction unit is used to predict pedestrian behaviors such as crossing the road and waiting based on characteristics such as pedestrian gait, stopping position, and line of sight, combined with the status of traffic lights at intersections, and generate early warning instructions in advance.

[0080] The vehicle trajectory prediction unit is used to predict whether a vehicle intends to change lanes illegally, speed, or brake suddenly by analyzing data such as vehicle speed, steering angle, and lane position, combined with information such as road speed limits and curve curvature.

[0081] The non-motorized vehicle situation prediction unit is used to predict risky behaviors such as driving against traffic, running red lights, and occupying motorized vehicle lanes by integrating geomagnetic detectors and video features for non-motorized vehicles such as electric vehicles and bicycles, thus filling the gap in non-motorized vehicle perception and prediction.

[0082] Please see Figure 5 The end-to-end security protection module includes:

[0083] The hardware anti-tampering unit integrates a Physically Unclonable Function (PUF) chip and a national cryptographic SM4 algorithm security chip (HSM) to generate a unique device identity and prevent chip replacement and firmware rewriting attacks.

[0084] The transmission encryption unit is used to encrypt transmission using a 5G-APN leased line and to perform end-to-end encryption on sensed data and control commands to prevent man-in-the-middle attacks and data sniffing.

[0085] The abnormal behavior monitoring unit is used to monitor the frequency and content of data transmission in real time, so as to immediately trigger a security alarm when abnormal data such as virtual obstacle forgery or traffic light status tampering is detected.

[0086] The security degradation unit is used to automatically switch to local offline working mode when encountering a severe attack, so as to provide a minimum early warning service based on pre-stored basic road information and avoid complete system paralysis.

[0087] Please see Figure 6 The dual-mode collaborative communication module includes:

[0088] The cross-protocol adapter unit is used to build a protocol conversion engine, which is compatible with mainstream protocols such as C-V2X, DSRC, and ETSI, so as to automatically identify the protocol type of the connected device and complete the real-time conversion.

[0089] The dual-mode switching unit is used to integrate 5G-V2X and Beidou short message communication modules to use 5G communication in urban areas with good signal and automatically switch to Beidou short message mode in signal blind spots such as tunnels and remote road sections.

[0090] The low-latency transmission unit is used to shorten the data transmission path by adopting edge node proximity access technology, and control the end-to-end communication latency to within 50 milliseconds to meet the needs of emergency early warning.

[0091] Please see Figure 7 The dynamic decision-making and scheduling module includes:

[0092] The scenario-based decision-making unit is used to preset various scenario models such as intersection traffic, highway hazard avoidance, and construction warning, so as to match the optimal decision-making scheme according to real-time traffic conditions and generate decision-making instructions such as vehicle speed suggestion, lane switching, and emergency braking.

[0093] The multi-object coordination unit is used to simultaneously send coordination instructions to the vehicle terminal, traffic control platform and surrounding RSUs to realize the three-way linkage of vehicle-road-platform. For example, when a traffic accident occurs, it can simultaneously notify the vehicles behind to slow down and inform the control platform to dispatch police forces.

[0094] Please see Figure 8 The green energy-saving adaptive module includes:

[0095] The dynamic energy consumption adjustment unit is used to adjust the sensor operating frequency according to the traffic flow density, reduce the camera frame rate and radar detection power during periods of sparse traffic, and adjust the transmission power of the dual-mode communication module according to the communication load to avoid unnecessary energy consumption.

[0096] The clean energy adaptation unit is used to adapt to outdoor clean energy power supply scenarios such as solar and wind power, and monitor the power supply stability in real time. When clean energy is sufficient, it will prioritize switching to clean energy power supply, and automatically switch to the mains power when it is insufficient, thereby reducing energy consumption costs.

[0097] The energy consumption monitoring and feedback unit is used to record the energy consumption data of each module, generate energy consumption analysis reports, and synchronize them to the full life cycle maintenance module to help determine whether the equipment is aging due to abnormal energy consumption, thus forming a collaborative closed loop of energy saving and maintenance.

[0098] Please see Figure 9 The full lifecycle maintenance module includes:

[0099] The real-time status monitoring unit is used to continuously monitor the working status of hardware such as sensors, communication modules, and chips, and record equipment operating parameters and energy consumption data;

[0100] The fault prediction and early warning unit is used to analyze the aging trend of equipment based on big data, and send early warning information to the operation and maintenance platform when the performance of components degrades to a threshold, so as to arrange maintenance in advance.

[0101] The remote upgrade unit supports remote OTA upgrades of firmware and algorithms without requiring on-site device disassembly, reducing maintenance costs and continuously optimizing terminal performance.

[0102] In practical use, the specific steps are as follows:

[0103] S1 integrates multiple types of sensors to simultaneously collect road information in all dimensions. It also adapts to severe weather and complex scenarios through environmental adaptive calibration and removes invalid data through target feature purification, providing accurate and comprehensive raw data support for subsequent processing.

[0104] S2 uses lightweight federated learning technology to balance privacy protection and distributed computing needs. It solves the problem of multi-source data synchronization through real-time data cleaning and is equipped with a dynamic priority scheduling mechanism to prioritize the processing of urgent event data, greatly reducing processing latency and ensuring the efficiency and consistency of data processing.

[0105] S3 focuses on three core traffic participants: pedestrians, vehicles, and non-motorized vehicles. It mines their key information, predicts risky behaviors, and generates early warning evidence, upgrading decision-making from passive response to proactive prediction.

[0106] S4 constructs a multi-layered protection system from hardware to data, relying on dedicated chips to achieve hardware anti-tampering, and using dedicated line encryption to ensure transmission security. It also monitors abnormal behavior in real time to resist attacks and is equipped with a security degradation mechanism to maintain core services in extreme situations, thus comprehensively strengthening the terminal security defense.

[0107] The S5 features a built-in cross-protocol adaptation engine that is compatible with mainstream communication standards. It also integrates a dual-mode communication module to handle signal differences in different scenarios and is equipped with low-latency transmission technology to achieve stable device connection and efficient information flow across regions and scenarios.

[0108] S6, based on a preset scene model, matches real-time road conditions and generates precise instructions. At the same time, it links the vehicle terminal, traffic control platform and surrounding RSUs to achieve a three-way coordinated response of vehicle-road-platform, quickly handle various traffic scenarios, and improve traffic efficiency and safety.

[0109] S7 dynamically adjusts equipment operating parameters based on traffic density and communication load, adapts to clean energy to achieve flexible power supply switching, and synchronously monitors the energy consumption data of each module to generate analysis reports, providing support for the maintenance module, and balancing operating efficiency and energy saving economy.

[0110] The S8 monitors the hardware operating status and parameters of each module in real time, and predicts the risk of device aging and failure through big data analysis, issuing early warnings. It also supports remote firmware upgrades, reducing manual troubleshooting costs, ensuring continuous and stable operation of the terminal, and achieving dynamic performance optimization.

[0111] Although the present invention has been described above with reference to embodiments, various modifications can be made and components can be replaced with equivalents without departing from the scope of the invention. In particular, as long as there is no structural conflict, the features in the disclosed embodiments can be combined with each other in any manner. The lack of an exhaustive description of these combinations in this specification is merely for the sake of brevity and resource conservation. Therefore, the present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A road side unit (RSU) terminal based on real-time information of a field road, comprising: include: The multi-source fusion perception module is used to integrate multiple types of sensors to simultaneously collect road information in all dimensions. It adapts to severe weather and complex scenarios through environmental adaptive calibration and removes invalid data through target feature purification. The intelligent edge processing module is used to process the data collected by the multi-source fusion sensing module through federated learning lightweighting and real-time data cleaning, and is equipped with a dynamic priority scheduling mechanism to prioritize the processing of emergency event data. The traffic participant intent prediction module is used to mine key information about pedestrians, vehicles, and non-motorized vehicles based on the data processed by the intelligent edge processing module, and to predict risky behaviors. The end-to-end security protection module is used to build a multi-layered protection system from hardware to data after the traffic participant intent prediction module predicts the traffic participant intent. It relies on a dedicated chip to achieve hardware anti-tampering, and uses dedicated line encryption to ensure transmission security. It also monitors abnormal behavior in real time to resist attacks, and is equipped with a security degradation mechanism. The dual-mode collaborative communication module is used to provide security protection after the full-link security protection module is implemented. It has a built-in cross-protocol adaptation engine to be compatible with mainstream communication standards, and integrates a dual-mode communication module to deal with signal differences in different scenarios, and is equipped with low-latency transmission technology. The dynamic decision-making and scheduling module is used to generate precise instructions by matching real-time traffic conditions with a preset scenario model after transmitting data using the dual-mode collaborative communication module, while simultaneously linking the vehicle terminal, traffic control platform and surrounding RSUs. The green energy-saving adaptive module is used to dynamically adjust the equipment operating parameters based on traffic density and communication load according to the scheduling data of the dynamic decision scheduling module, and adapt to clean energy to achieve flexible power supply switching, and simultaneously monitor the energy consumption data of each module to generate analysis reports. The full lifecycle maintenance module is used to monitor the hardware operating status and parameters of each module in real time, and predict the risk of equipment aging and failure through big data analysis, and issue early warnings. It also supports remote firmware upgrades. The end-to-end security protection module includes: The hardware anti-tampering unit integrates a physically unclonable function chip and a national cryptographic SM4 algorithm security chip to generate a unique device identity. The transmission encryption unit is used for encrypted transmission using a 5G-APN leased line and for end-to-end encryption of sensing data and control commands. An abnormal behavior monitoring unit is used to monitor the frequency and content of data transmission in real time. The security degradation unit is used to automatically switch to local offline working mode in the event of a severe attack. 2.The RSU terminal based on real-time information of a live road according to claim 1, wherein, The multi-source fusion sensing module includes: Multi-source heterogeneous data acquisition unit, used to simultaneously collect all dimensions of road data; An environmental adaptive calibration unit is used to monitor weather and lighting conditions in real time to automatically switch the sensor's operating mode in adverse weather conditions. The target feature purification unit is used to initially filter data through edge computing and retain the accurate feature data of key targets. 3.The RSU terminal based on real-time information of a live road according to claim 1, characterized in that, The intelligent edge processing module includes: The Federated Learning Lightweight Unit is used to employ the improved FTD3 algorithm, transmitting only neural network parameters instead of the original data, and enabling distributed computing across RSU clusters. The real-time data cleaning unit is used to control the time difference between different sensors within a threshold through outlier removal and timestamp alignment algorithms. The dynamic priority scheduling unit is used to classify data processing levels according to the urgency of events. 4.The RSU terminal based on real-time information of a live road according to claim 1, wherein, The traffic participant intention prediction module includes: The pedestrian behavior prediction unit is used to predict pedestrian behavior based on pedestrian characteristics and the status of traffic lights at intersections. The vehicle trajectory prediction unit is used to predict vehicle intentions by analyzing vehicle data and combining it with road information. The non-motorized vehicle situation prediction unit is used to predict the risky behavior of non-motorized vehicles by integrating geomagnetic detectors and video features. 5.The RSU terminal based on real-time information of a live road according to claim 1, characterized in that, The dual-mode collaborative communication module includes: Cross-protocol adaptation unit, used for built-in protocol conversion engine, compatible with mainstream protocols, to automatically identify the protocol type of the connected device and complete real-time conversion; The dual-mode switching unit is used to integrate 5G-V2X and Beidou short message communication modules to use 5G communication in areas with good signal and automatically switch to Beidou short message mode in areas with no signal. Low-latency transmission units are used to shorten data transmission paths by employing edge node proximity access technology. 6.The RSU terminal based on real-time information of a live road according to claim 1, wherein, The dynamic decision-making and scheduling module includes: The scenario-based decision-making unit is used to preset multiple scenario models to match the optimal decision-making scheme based on real-time traffic conditions and generate decision instructions; The multi-object coordination unit is used to simultaneously send coordination instructions to the vehicle terminal, traffic control platform and surrounding RSUs.

7. A vehicle-road cooperative RSU terminal based on real-time on-site road information according to claim 1, characterized in that, The green energy-saving adaptive module includes: The energy consumption dynamic adjustment unit is used to adjust the sensor operating frequency according to traffic flow density and adjust the transmission power of the dual-mode communication module according to communication load. The clean energy adapter unit is used to adapt to outdoor clean energy power supply scenarios, monitor the power supply stability in real time, and switch to clean energy power supply first when clean energy is sufficient, and automatically switch to mains power when it is insufficient, thereby reducing energy consumption costs. The energy consumption monitoring and feedback unit is used to record the energy consumption data of each module, generate energy consumption analysis reports, and synchronize them to the full life cycle maintenance module.

8. A vehicle-road cooperative RSU terminal based on real-time on-site road information according to claim 1, characterized in that, The full lifecycle maintenance module includes: The real-time status monitoring unit is used to continuously monitor the working status of the hardware and record the device's operating parameters and energy consumption data; The fault prediction and early warning unit is used to analyze the aging trend of equipment based on big data, and send early warning information to the operation and maintenance platform when the performance of components degrades to a threshold. The remote upgrade unit is used to support remote OTA upgrades of firmware and algorithms.

Images