Low-speed traffic participant and vehicle collaborative anti-collision early warning system based on cellular network

Abstract

The invention discloses a low-speed traffic participant and vehicle collaborative anti-collision early warning system based on a cellular network, which belongs to the technical field of intelligent networked automobiles and comprises a multi-terminal real-time positioning and communication module, a collision risk operation and edge optimization module, a graded early warning and privacy protection module and a dynamic scene adaptation module. The multi-terminal real-time positioning and communication module is in wireless connection with the collision risk operation and edge optimization module, the collision risk operation and edge optimization module is in wireless connection with the grading early warning and privacy protection module, and the dynamic scene adaptation module is in wireless connection with the multi-terminal real-time positioning and communication module. Through communication technology multiplexing, algorithm lightweight and privacy protection design, a low-cost and high-coverage real-time anti-collision early warning system is constructed, active safety protection of global traffic participants is achieved at the lowest cost, and particularly the blank that non-motor vehicles and pedestrians lack intelligent protection means is made up.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent connected vehicles, and specifically to a collaborative anti-collision warning system for low-speed traffic participants and vehicles based on a cellular network. Background Art

[0002] ITS integrates advanced information technology and data communication transmission technologies to achieve intelligent traffic management. V2X technology covers vehicle-to-vehicle (V2V), vehicle-to-pedestrian (V2P), and vehicle-to-infrastructure (V2I) communications. The combination of the two focuses on safety protection and real-time communication and positioning technology between motor vehicles, non-motor vehicles, and pedestrians, effectively reducing the risk of traffic accidents and improving travel safety and traffic efficiency.

[0003] However, due to the obstruction of road greening or buildings, the driving environment will inevitably have a reduced field of vision or even blind spots. The existence of road blind spots greatly limits the driving vision of human drivers and reduces the perception distance of intelligent connected vehicles, which in turn brings higher driving safety risks. Especially in low-speed traffic or intersection environments, collision accidents are very likely to occur between vehicles in different directions in the blind spots. For this reason, a low-speed traffic participant and vehicle collaborative anti-collision warning system based on cellular networks is proposed. Summary of the Invention

[0004] The purpose of the present invention is to provide a cellular network-based low-speed traffic participant and vehicle collaborative collision warning system to solve the problems raised in the above background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solutions: a cellular network-based low-speed traffic participant and vehicle collaborative collision avoidance warning system, comprising a multi-terminal real-time positioning and communication module, a collision risk calculation and edge optimization module, a hierarchical warning and privacy protection module, and a dynamic scene adaptation module, wherein the multi-terminal real-time positioning and communication module is wirelessly connected to the collision risk calculation and edge optimization module, the collision risk calculation and edge optimization module is wirelessly connected to the hierarchical warning and privacy protection module, and the dynamic scene adaptation module is wirelessly connected to the multi-terminal real-time positioning and communication module;

[0006] The multi-terminal real-time positioning and communication module includes a positioning technology combination unit and a communication protocol setting unit. The positioning technology combination unit includes a mobile phone terminal and a vehicle terminal. The communication protocol setting unit includes a data transmission layer and a dynamic switching of communication modes.

[0007] The mobile phone includes satellite positioning, base station-assisted positioning, and sensor fusion. The satellite positioning uses GPS or Beidou dual-mode, with an accuracy of 1m-3m in open areas. The base station-assisted positioning compensates for satellite signal obstruction through cellular network triangulation. The sensor fusion uses the phone's gyroscope or accelerometer data to assist in trajectory prediction.

[0008] Among them, cellular network triangulation uses TDOA algorithm;

[0009] The vehicle side includes high-precision positioning and redundancy design. The high-precision positioning uses dual-frequency GNSS + vehicle CAN bus speed data. The redundancy design is used for temporary continuation of inertial navigation calculation when satellite signals are lost.

[0010] Among them, the dual-frequency GNSS uses u-blox ZED-F9P with an accuracy of 0.1m.

[0011] Among them, the data transmission layer adopts the lightweight MQTT protocol to support high-frequency small data packet transmission under low bandwidth, and the base station side deploys HTTP / 2 long connection to reduce handshake delay. The dynamic switching of the communication mode includes normal scenarios and weak signal scenarios. The normal scenario adopts 5G cellular network with the highest priority and delay <50ms. In the weak signal scenario, it automatically downgrades to 4G cellular network or switches to Wi-Fi connection or Bluetooth connection.

[0012] As a further preferred embodiment of the present technical solution: the collision risk calculation and edge optimization module includes an algorithm operation unit and an edge algorithm optimization unit, the algorithm operation unit includes a core algorithm and trajectory prediction, the core algorithm defines a function named ttc-with-error for calculating a time collision model of a comprehensive positioning error, wherein the specific operation mode of the time collision model includes the following steps:

[0013] A1. Define a function that accepts multiple parameters.

[0014] A2. Calculate the relative speed of vehicles and pedestrians;

[0015] A3. Use the linalg.norm function in the numpy library to calculate the distance between the vehicle and the pedestrian.

[0016] A4. Comprehensive positioning error, calculate the worst-case time collision model;

[0017] A5. The function returns a tuple including ttc-min and ttc-max.

[0018] As a further preferred embodiment of the present technical solution: in A1, the multiple parameters include the position of the vehicle, the speed of the vehicle, the position of the pedestrian, the speed of the pedestrian and the comprehensive positioning error, and the default value of the comprehensive positioning error is set to 1.5;

[0019] In A2, the calculation method is vehicle speed minus pedestrian speed;

[0020] In A4, when the relative speed is greater than 0, ttc-min is the distance minus the error range divided by the relative speed, and ttc-max is the distance plus the error range divided by the relative speed. When the relative speed is less than 0, both ttc-min and ttc-max are set to infinity.

[0021] As a further preferred embodiment of the present technical solution: the trajectory prediction is based on a kinematic model, and the mobile phone predicts the trajectory in the next 2 seconds through the mobile phone sensor data.

[0022] As a further preferred embodiment of the present technical solution: the edge algorithm optimization unit includes distributed computing nodes, dynamic load balancing and data filtering. The distributed computing nodes deploy edge servers on the base station side and process collision calculations in accordance with the proximity principle, thereby reducing round-trip delays in the cloud. The dynamic load balancing allocates computing resources according to regional traffic density. The data filtering only performs collision judgment on traffic participants within a radius of 100m, thereby reducing invalid calculations.

[0023] As a further preferred embodiment of the present technical solution: the hierarchical warning and privacy protection module includes a warning mechanism triggering unit and a privacy protection setting unit, and the warning mechanism triggering unit includes a hierarchical strategy and a vehicle control interface;

[0024] The grading strategy includes at least three risk levels, namely low risk warning, medium risk warning, and high risk warning. The three risk levels are determined based on the range of the time collision model. After matching the corresponding risk level, both the vehicle and the mobile phone initiate an adaptive response.

[0025] The vehicle control interface sends speed limit instructions via the CAN bus and is linked to the ADAS system.

[0026] As a further preferred embodiment of the present technical solution: the privacy protection setting unit includes data anonymization and data lifecycle management. For data anonymization, the location data is associated with a temporary random ID, which is changed every 5 minutes and transmitted using HMAC-SHA256 encryption to prevent man-in-the-middle attacks. For data lifecycle management, the original data is retained at the edge node for no more than 10 seconds and is deleted immediately after the calculation is completed, and only the desensitized statistical results are saved.

[0027] As a further preferred embodiment of the present technical solution: the dynamic scene adaptation module is used to compensate for the multi-terminal real-time positioning and communication module, the dynamic scene adaptation module includes positioning frequency adjustment and communication mode switching, the positioning frequency adjustment includes congested areas and open roads, and the communication mode switching includes signal strength threshold and short-range communication priority.

[0028] As a further preferred embodiment of the present technical solution: the congested area is positioned at a high frequency of 1 Hz, once per second, and the open road is positioned at a low frequency of 0.5 Hz, once every two seconds;

[0029] Among them, when the signal strength threshold is 5G cellular network signal <-90dBm, it switches to 4G cellular network signal. When the short-range communication priority is when the distance between the vehicle and the pedestrian is <10m, Bluetooth direct connection is started to avoid base station transfer delay.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] 1. This invention builds a low-cost, high-coverage, real-time collision avoidance and warning system through communication technology reuse, algorithm lightweighting, and privacy protection design. It aims to achieve active safety protection for all traffic participants at the lowest cost. In particular, it fills the gap in intelligent protection for non-motorized vehicles (low-speed electric two-wheelers, three-wheelers, and four-wheelers) and pedestrians. By reusing existing cellular networks and mobile phone terminals, it enriches the application scope of V2X technology to include low-speed traffic participants such as non-motorized vehicles. The core technology revolves around the positioning-computing-warning chain, combining practicality and innovation.

[0032] 2. This invention is not only a technological breakthrough, but also aims to upgrade the paradigm of traffic safety from "single-vehicle intelligence" to "group collaboration." By sharing the real-time location of mobile phones and vehicles, potential collision risks can be predicted in advance, triggering warnings even if the target is in the sensor's blind spot, thereby improving the safety of vulnerable road users. This effectively solves the problem that traditional vehicle collision avoidance systems (such as automatic emergency braking) rely on on-board sensors (radar, cameras) but cannot detect pedestrians or electric vehicles outside the visual blind spot (for example, in the "ghost peeking" scenario when the vehicle turns right).

[0033] 3. This invention reuses mobile phones and cellular networks. Pedestrians or electric vehicles do not need additional hardware and only need to install a lightweight app. Edge computing replaces roadside equipment and uses existing edge nodes on the base station side to process data, reducing infrastructure construction investment, thereby lowering the cost and threshold of technology deployment. This effectively solves the problem that existing V2X solutions require vehicles to install dedicated OBUs (on-board units) and roadside equipment (RSUs), which are costly and difficult to cover non-motor vehicles and pedestrians.

[0034] 4. The present invention uses multi-communication mode collaboration, through cellular network (4G or 5G) wide-area coverage + Wi-Fi or Bluetooth short-range blind spot filling (such as parking lot scenarios), and dynamic positioning frequency adjustment. The positioning frequency is increased (1Hz) in congested areas and reduced (0.5Hz) in open areas, balancing accuracy and energy consumption, thereby achieving full-scene coverage and real-time response. This effectively solves the problem that a single communication mode (such as pure C-V2X) is easily affected by signal obstruction in complex urban environments (such as underground garages and tunnels), resulting in warning failure.

[0035] 5. The present invention uses anonymization processing to associate location information with a randomly generated temporary ID, rather than a real identity. Furthermore, by utilizing the principle of data minimization, only the data required for real-time calculations is retained, and historical tracks are automatically deleted after 10 seconds. This effectively protects user privacy and data security, thereby avoiding the problem of direct transmission of location data that could leak user whereabouts and cause privacy risks.

[0036] 6. This invention has low cost and high coverage, reuses mobile phones and cellular networks, and does not require the modification of non-motor vehicles or roadside facilities. Through the complementarity of multiple communication modes, it adapts to complex environments such as underground and tunnels, enhancing the overall robustness of the system in all scenarios. Edge computing + TTC algorithm optimization controls the delay within 200ms, greatly enhancing the overall real-time performance and accuracy of the system. At the same time, anonymization + encrypted transmission complies with data protection regulations such as GDPR, effectively ensuring the privacy and security of the overall operation of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is a schematic diagram of the architecture of the cellular network-based low-speed traffic participant and vehicle collaborative collision warning system of the present invention;

[0038] Figure 2 A schematic flow chart of the collision risk calculation and edge optimization module in the cellular network-based low-speed traffic participant and vehicle collaborative collision avoidance warning system of the present invention;

[0039] Figure 3 The figure is a flow chart of a time collision model in the collaborative anti-collision warning system for low-speed traffic participants and vehicles based on a cellular network of the present invention. DETAILED DESCRIPTION

[0040] The technical solutions in the embodiments of the present invention will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present invention.

[0041] Example

[0042] See also Figure 1-Figure 3The present invention provides a technical solution: a collaborative collision avoidance warning system for low-speed traffic participants and vehicles based on a cellular network, comprising a multi-terminal real-time positioning and communication module, a collision risk calculation and edge optimization module, a hierarchical warning and privacy protection module, and a dynamic scene adaptation module. The multi-terminal real-time positioning and communication module is wirelessly connected to the collision risk calculation and edge optimization module, the collision risk calculation and edge optimization module is wirelessly connected to the hierarchical warning and privacy protection module, and the dynamic scene adaptation module is wirelessly connected to the multi-terminal real-time positioning and communication module.

[0043] The multi-terminal real-time positioning and communication module includes a positioning technology combination unit and a communication protocol setting unit. The positioning technology combination unit includes a mobile phone terminal and a vehicle terminal. The communication protocol setting unit includes a data transmission layer and a dynamic switching of communication modes.

[0044] The mobile phone side includes satellite positioning, base station-assisted positioning, and sensor fusion. Satellite positioning uses GPS or Beidou dual-mode, with an accuracy of 1m-3m in open areas. Base station-assisted positioning compensates for satellite signal obstruction scenarios through cellular network triangulation, which uses the TDOA algorithm. Sensor fusion uses the phone's gyroscope or accelerometer data to assist in trajectory prediction.

[0045] The vehicle side includes high-precision positioning and redundancy design. High-precision positioning uses dual-frequency GNSS + vehicle CAN bus speed data. The dual-frequency GNSS uses u-blox ZED-F9P with an accuracy of 0.1m. The redundancy design is used for the inertial navigation unit (IMU) to temporarily continue calculation when the satellite signal is lost.

[0046] Among them, the data transmission layer adopts the lightweight MQTT protocol to support high-frequency small data packet transmission (position coordinates + velocity vector) under low bandwidth, and HTTP / 2 long connection is deployed on the base station side to reduce handshake delay. The communication mode is dynamically switched including normal scenarios and weak signal scenarios. In the normal scenario, the 5G cellular network is used with the highest priority and delay <50ms. In the weak signal scenario, it automatically downgrades to the 4G cellular network or switches to Wi-Fi connection or Bluetooth connection.

[0047] In this embodiment, specifically, the collision risk calculation and edge optimization module includes an algorithm operation unit and an edge algorithm optimization unit. The algorithm operation unit includes a core algorithm and trajectory prediction. The core algorithm defines a function called ttc-with-error, which is used to calculate the time collision model (TTC) of the comprehensive positioning error. The specific operation mode of the time collision model (TTC) includes the following steps:

[0048] A1. Define a function that accepts multiple parameters.

[0049] A2. Calculate the relative speed of vehicles and pedestrians;

[0050] A3. Use the linalg.norm function in the numpy library to calculate the distance between the vehicle and the pedestrian.

[0051] A4. Calculate the worst-case time-to-collision model (TTC) based on the integrated positioning error.

[0052] A5. The function returns a tuple including ttc-min and ttc-max.

[0053] In this embodiment, specifically: in A1, the multiple parameters include the vehicle position pos-car, the vehicle speed vel-car, the pedestrian position pos-ped, the pedestrian speed vel-car and the comprehensive positioning error error-range, and the default value of the comprehensive positioning error error-range is set to 1.5;

[0054] In A2, the calculation method is vehicle speed minus pedestrian speed;

[0055] In A4, when the relative speed is greater than 0, ttc-min is the distance minus the error range divided by the relative speed, and ttc-max is the distance plus the error range divided by the relative speed. When the relative speed is less than 0, both ttc-min and ttc-max are set to infinity (INF).

[0056] In this embodiment, specifically: trajectory prediction is based on a kinematic model (constant velocity or acceleration assumption), and the mobile phone predicts the trajectory for the next 5 seconds through mobile phone sensor data.

[0057] In this embodiment, specifically: the edge algorithm optimization unit includes distributed computing nodes, dynamic load balancing and data filtering. The distributed computing nodes deploy edge servers on the base station side, and process collision calculations in accordance with the proximity principle, thereby reducing round-trip delays in the cloud. Dynamic load balancing allocates computing resources according to regional traffic density, and data filtering only performs collision judgment on traffic participants within a radius of 100m, thereby reducing invalid calculations.

[0058] In this embodiment, specifically: the hierarchical warning and privacy protection module includes a warning mechanism trigger unit and a privacy protection setting unit, and the warning mechanism trigger unit includes a hierarchical strategy and a vehicle control interface;

[0059] The grading strategy includes at least three risk levels: low risk warning, medium risk warning, and high risk warning. The three risk levels are determined based on the range of the time collision model. After matching the corresponding risk level, both the vehicle and the mobile phone initiate an adaptive response, as shown in the grading strategy in Table 1 below:

[0060] Risk Level TTC range Vehicle-side response Mobile phone response Low risk warning 5s<TTC≤10s The central control screen displays a yellow warning icon APP slight vibration Medium risk warning 3s<TTC≤5s The buzzer sounds intermittently Screen flickering (50% brightness) High-risk warning TTC≤3s Buzzer continues to sound + automatic speed reduction Screen flickering with strong light (100% brightness)

[0061] Table 1: Classification strategy

[0062] The vehicle control interface sends speed limit instructions via the CAN bus and is linked to the ADAS system.

[0063] In this embodiment, specifically: the privacy protection setting unit includes data anonymization and data lifecycle management. For data anonymization, the location data is associated with a temporary random ID, which is changed every 5 minutes and transmitted using HMAC-SHA256 encryption to prevent man-in-the-middle attacks. For data lifecycle management, the original data is retained at the edge node for no more than 10 seconds and is deleted immediately after the calculation is completed, and only the desensitized statistical results are saved.

[0064] In this embodiment, specifically: the dynamic scene adaptation module is used to compensate for the multi-terminal real-time positioning and communication module, the dynamic scene adaptation module includes positioning frequency adjustment and communication mode switching, the positioning frequency adjustment includes congested areas and open roads, and the communication mode switching includes signal strength threshold and short-range communication priority.

[0065] In this embodiment, specifically: congested areas use 1 Hz high-frequency positioning, once per second, and open roads use 0.5 Hz low-frequency positioning, once every two seconds;

[0066] Among them, when the signal strength threshold of the 5G cellular network signal is less than -90dBm, it switches to the 4G cellular network signal. When the distance between the vehicle and the pedestrian is less than 10m, the short-range communication priority is to start the Bluetooth direct connection to avoid the base station transfer delay.

[0067] Working Principle: The mobile phone integrates a GPS or Beidou positioning module (accuracy of 1-3 meters) and a 4G / 5G communication module, and implements data encryption and anonymized transmission through the SDK. The vehicle terminal includes a 5G communication module, a high-precision dual-frequency GNSS positioning module (accuracy of 0.1 meters), and a CAN bus interface (for obtaining vehicle speed data).

[0068] Cellular base stations (4G or 5G) serve as the data transmission backbone, and edge computing nodes are deployed on the base station side to achieve local data processing (i.e., TTC calculation). In weak signal areas (such as underground parking lots), Wi-Fi or Bluetooth direct connections are used to fill in blind spots and reduce communication delays.

[0069] Among them, mobile phones and vehicles upload real-time location data (frequency is adjustable: 1Hz in congested areas, 0.5Hz in open areas) to edge nodes. After the edge nodes perform TTC calculations, they push the risk level to the vehicle and mobile phone terminals, triggering graded warnings. The specific operations are as follows:

[0070] The real-time position and velocity vectors of the mobile phone (pedestrian or non-motorized vehicle) and the vehicle are input, and the Kalman filter algorithm is used to reduce positioning jitter and improve trajectory prediction accuracy, thereby performing error compensation. At the same time, based on the TTC (Time to Collision) model and combined with the kinematic equation, the trajectory of the next 5 seconds is predicted to achieve risk assessment.

[0071] Finally, the cloud platform is responsible for long-term data storage (such as accident records) and global traffic flow analysis, supporting the dynamic optimization of multi-terminal collaborative warning strategies.

[0072] Note: Further elaboration is needed on the following: data cleaning filters out positioning anomalies (such as satellite signal jumps), error compensation introduces a positioning error range (such as ±1.5 meters), and calculates the worst-case TTC boundary value.

[0073] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A cellular network-based, low-speed traffic participant and vehicle collaborative collision warning system, including a multi-terminal real-time positioning and communication module, a collision risk calculation and edge optimization module, a hierarchical warning and privacy protection module, and a dynamic scene adaptation module. Its features include: The multi-terminal real-time positioning and communication module is wirelessly connected to the collision risk calculation and edge optimization module, the collision risk calculation and edge optimization module is wirelessly connected to the hierarchical warning and privacy protection module, and the dynamic scene adaptation module is wirelessly connected to the multi-terminal real-time positioning and communication module; The multi-terminal real-time positioning and communication module includes a positioning technology combination unit and a communication protocol setting unit. The positioning technology combination unit includes a mobile phone terminal and a vehicle terminal. The communication protocol setting unit includes a data transmission layer and a dynamic switching of communication modes. The mobile phone includes satellite positioning, base station-assisted positioning, and sensor fusion. The satellite positioning uses GPS or Beidou dual-mode, with an accuracy of 1m-3m in open areas. The base station-assisted positioning compensates for satellite signal obstruction through cellular network triangulation. The sensor fusion uses the phone's gyroscope or accelerometer data to assist in trajectory prediction. The vehicle side includes high-precision positioning and redundancy design. The high-precision positioning uses dual-frequency GNSS + vehicle CAN bus speed data. The redundancy design is used for temporary continuation of inertial navigation calculation when satellite signals are lost. Among them, the data transmission layer adopts the lightweight MQTT protocol to support high-frequency small data packet transmission under low bandwidth, and the base station side deploys HTTP / 2 long connection to reduce handshake delay. The dynamic switching of the communication mode includes normal scenarios and weak signal scenarios. The normal scenario adopts 5G cellular network with the highest priority and delay <50ms. In the weak signal scenario, it automatically downgrades to 4G cellular network or switches to Wi-Fi connection or Bluetooth connection.

2. The cellular network-based low-speed traffic participant and vehicle collaborative collision warning system according to claim 1, characterized in that: The collision risk calculation and edge optimization module includes an algorithm operation unit and an edge algorithm optimization unit. The algorithm operation unit includes a core algorithm and trajectory prediction. The core algorithm defines a function called ttc-with-error for calculating a time collision model of a comprehensive positioning error. The specific operation mode of the time collision model includes the following steps: A1. Define a function that accepts multiple parameters. A2. Calculate the relative speed of vehicles and pedestrians; A3. Use the linalg.norm function in the numpy library to calculate the distance between the vehicle and the pedestrian. A4. Comprehensive positioning error, calculate the worst-case time collision model; A5. The function returns a tuple including ttc-min and ttc-max.

3. The cellular network-based low-speed traffic participant and vehicle collaborative collision warning system according to claim 2, characterized in that: In A1, multiple parameters include vehicle position, vehicle speed, pedestrian position, pedestrian speed, and comprehensive positioning error, and the default value of the comprehensive positioning error is set to 1.5; In A2, the calculation method is vehicle speed minus pedestrian speed; In A4, when the relative speed is greater than 0, ttc-min is the distance minus the error range divided by the relative speed, and ttc-max is the distance plus the error range divided by the relative speed. When the relative speed is less than 0, both ttc-min and ttc-max are set to infinity.

4. The cellular network-based low-speed traffic participant and vehicle collaborative collision warning system according to claim 2, characterized in that: The edge algorithm optimization unit includes distributed computing nodes, dynamic load balancing and data filtering. The distributed computing nodes deploy edge servers on the base station side and process collision calculations based on the proximity principle. The dynamic load balancing allocates computing resources according to regional traffic density. The data filtering only performs collision judgment on traffic participants within a radius of 100m.

5. The cellular network-based low-speed traffic participant and vehicle collaborative collision warning system according to claim 2, characterized in that: The hierarchical warning and privacy protection module includes a warning mechanism triggering unit and a privacy protection setting unit. The warning mechanism triggering unit includes a hierarchical strategy and a vehicle control interface. The grading strategy includes at least three risk levels, namely low risk warning, medium risk warning, and high risk warning. The three risk levels are determined based on the range of the time collision model. After matching the corresponding risk level, both the vehicle and the mobile phone initiate an adaptive response. The vehicle control interface sends speed limit instructions via the CAN bus and is linked to the ADAS system.

6. The cellular network-based low-speed traffic participant and vehicle collaborative collision warning system according to claim 5, characterized in that: The privacy protection setting unit includes data anonymization and data lifecycle management. For data anonymization, the location data is associated with a temporary random ID and transmitted using HMAC-SHA256 encryption. For data lifecycle management, the original data is retained at the edge node for no more than 10 seconds and is deleted immediately after the calculation is completed, and only the desensitized statistical results are saved.

7. The cellular network-based low-speed traffic participant and vehicle collaborative collision warning system according to claim 1, characterized in that: The dynamic scene adaptation module is used to compensate for the multi-terminal real-time positioning and communication module. The dynamic scene adaptation module includes positioning frequency adjustment and communication mode switching. The positioning frequency adjustment includes congested areas and open roads, and the communication mode switching includes signal strength threshold and short-range communication priority.

8. The cellular network-based low-speed traffic participant and vehicle collaborative collision warning system according to claim 7, characterized in that: The congested area is positioned at a high frequency of 1 Hz, once per second, and the open road is positioned at a low frequency of 0.5 Hz, once every two seconds; Among them, when the signal strength threshold is 5G cellular network signal <-90dBm, it switches to 4G cellular network signal, and when the short-range communication priority is the distance between the vehicle and the pedestrian is <10m, Bluetooth direct connection is started.

Images