Control method and device of vehicle, vehicle and electronic equipment

Abstract

Embodiments of the present application provide a vehicle control method and device, a vehicle and an electronic device. The method comprises: in response to vehicle networking data indicating that a risk event exists for the vehicle, obtaining visibility data of an environment in which the vehicle is currently located by using vehicle networking technology; determining a weight coefficient corresponding to each of a plurality of sensors of the vehicle based on the visibility data; performing weighted fusion processing on perception data of the plurality of sensors based on the weight coefficient corresponding to each of the plurality of sensors to obtain a fusion result; verifying the risk event based on the fusion result to obtain a verification result; in response to the verification result indicating that a target road object in the risk event exists with a collision risk for the vehicle, generating a control instruction based on a predicted collision time between the vehicle and the target road object; and controlling the vehicle based on the control instruction. The present application solves the technical problem that a road object with a collision risk for the vehicle cannot be accurately identified in bad weather, resulting in low vehicle driving safety.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, and more particularly to a vehicle control method, device, vehicle, and electronic equipment. Background Technology

[0002] With the rapid development of vehicle intelligence and connectivity technologies, intelligent connected vehicles have become an important trend in the automotive industry. These vehicles are equipped with advanced intelligent driving functions, providing drivers with advanced driver assistance or enabling fully autonomous driving.

[0003] In related technologies, vehicles typically rely on multiple sensors to perceive their surroundings and road conditions in real time. Under favorable weather conditions, these sensors provide relatively accurate and reliable environmental perception data to ensure safe driving. However, in adverse weather conditions, sensor performance deteriorates significantly, making it difficult to accurately identify road objects that pose a collision risk to the vehicle, thus reducing driving safety.

[0004] There is currently no effective solution to the aforementioned technical problems that result in low vehicle driving safety. Summary of the Invention

[0005] This application provides a vehicle control method, device, vehicle, and electronic device, aiming to improve the technical problem of low vehicle driving safety caused by the inability to accurately identify road objects that pose a collision risk to the vehicle in adverse weather conditions.

[0006] According to one aspect of the embodiments of this application, a vehicle control method is provided. The method includes: in response to vehicle-to-everything (V2X) data indicating a risk event for the vehicle, acquiring visibility data of the vehicle's current environment via V2X technology; determining weight coefficients for multiple sensors of the vehicle based on the visibility data; performing weighted fusion processing on the perception data from the multiple sensors based on the weight coefficients to obtain a fusion result, wherein the perception data is used to characterize the road state of the road currently being traveled by the vehicle; verifying the risk event based on the fusion result to obtain a verification result; in response to the verification result indicating the presence of a target road object in the risk event that poses a collision risk to the vehicle, generating a control command based on the predicted collision time between the vehicle and the target road object; and controlling the vehicle based on the control command.

[0007] In the vehicle control method provided in this application, after determining that a risk event exists for the vehicle through vehicle-to-everything (V2X) data, visibility data of the vehicle's current environment can be further obtained through V2X technology. Based on the visibility data, the weighting coefficients of multiple sensors in the vehicle are dynamically adjusted to ensure the accuracy of the fusion result obtained when the perception data from multiple sensors is weighted and fused according to the weighting coefficients. Then, the risk event is further verified using the fusion result to determine whether there is a target road object that poses a collision risk to the vehicle. If so, a control command is generated based on the predicted collision time between the vehicle and the target road object; based on the control command, the vehicle is controlled. In other words, in this embodiment, dual detection using the fusion result of V2X data and the perception data from multiple vehicle sensors can improve the detection accuracy of target road objects that pose a collision risk to the vehicle, reduce false alarms and false negatives, and generate control commands based on the predicted collision time between the vehicle and the target road object to control the vehicle. This improves vehicle driving safety and solves the technical problem in related technologies where the inability to accurately identify road objects that pose a collision risk to the vehicle in adverse weather conditions leads to low vehicle driving safety.

[0008] Optionally, based on visibility data, the weight coefficients corresponding to multiple sensors of the vehicle are determined, including: based on visibility data, determining the weather level of the current environment in which the vehicle is located; based on the weather level, determining the weight coefficients corresponding to multiple sensors from a weighted weather level mapping table, wherein the weighted weather level mapping table includes mapping relationships between various weather levels and sensor weight coefficients.

[0009] In this embodiment, since different types of sensors exhibit different performance under different environmental conditions, the weather level of the current environment can be determined based on the visibility data of the vehicle's current location. Then, based on the weather level, the weight coefficients corresponding to multiple sensors in the vehicle are determined from the weighted weather level mapping table. This allows for dynamic adjustment of the weight coefficients of multiple sensors in the vehicle according to the weather level of the current environment, maximizing the advantages of different sensors under different environmental conditions. By performing weighted fusion processing on the sensing data of multiple sensors according to the weight coefficients, the shortcomings of a single sensor under specific environmental conditions can be compensated for, thereby improving the accuracy of the fusion result.

[0010] Optionally, based on the fusion result, the risk event is verified to obtain a verification result, including: based on the fusion result, identifying road objects located in front of the vehicle on the road currently being driven by the vehicle to obtain an identification result, wherein the identification result is used to characterize whether there is a road object in front of the vehicle and the distance between the road object and the vehicle; in response to the identification result indicating that there is a road object in front of the vehicle and the distance between the road object and the vehicle is less than a preset distance threshold, determining that the verification result indicates that there is a target road object in the risk event that poses a collision risk to the vehicle.

[0011] In this embodiment, since the fusion result includes perception data from multiple sensors, the fusion result can comprehensively and accurately identify road objects located in front of the vehicle's current driving road and the distance between the road object and the vehicle. If the identification result indicates that there is a road object in front of the vehicle whose distance to the vehicle is less than a preset distance threshold, then the road object is identified as a target road object that poses a collision risk to the vehicle. In this way, the identification of target road objects that pose a collision risk to the vehicle can be improved by using the fusion result of perception data from multiple sensors.

[0012] Optionally, a control command is generated based on the collision time between the vehicle and the target road object, including: determining the predicted collision time between the vehicle and the target road object based on the distance between them; generating a first control command in response to the predicted collision time being less than a preset duration, wherein the first control command is used to control the vehicle to brake suddenly; and generating a second control command in response to the predicted collision time being greater than or equal to the preset duration, wherein the second control command is used to control the vehicle to issue a warning.

[0013] In this embodiment, the predicted collision time between the vehicle and the target road object is used as the basis for vehicle control decisions. When the predicted collision time is less than a preset duration, the vehicle is controlled to brake urgently. When the predicted collision time is greater than or equal to the preset duration, the vehicle is controlled to issue an alarm. This hierarchical response mechanism based on the predicted collision time between the target road object and the vehicle can achieve a refined response to risk events. It can not only improve the vehicle's response speed and decision-making quality when facing collision risks, but also reduce misoperation and excessive intervention, optimize energy management, and improve driving experience and safety.

[0014] Optionally, in response to vehicle network data indicating a risk event in the vehicle, the method further includes: receiving vehicle network data through a dedicated communication channel in response to a risk level higher than a preset risk level threshold, wherein the dedicated communication channel is used to transmit communication data with a correlation to safe vehicle driving higher than the preset threshold; and receiving vehicle network data through a shared communication channel in response to a risk level lower than or equal to the preset risk level threshold, wherein the shared communication channel is used to transmit communication data with a correlation to safe vehicle driving lower than or equal to the preset threshold.

[0015] In this embodiment, by differentiating the risk levels of risk events, vehicle-to-everything (V2X) data is transmitted on different communication channels, achieving effective resource management and scheduling. V2X data carrying high-risk events is transmitted through a dedicated communication channel, ensuring that high-risk events are not interfered with by other non-urgent data, achieving ultra-low latency transmission, and enabling vehicles to receive high-risk events promptly and react quickly. Conversely, V2X data carrying low-risk events is transmitted through a shared communication channel. This ensures the normal flow of V2X data while avoiding the occupation of communication resources for high-risk events, achieving efficient utilization of communication resources.

[0016] Optionally, risk events are associated with risk levels.

[0017] In this embodiment, different risk levels are set for risk events, enabling a more refined assessment of potential hazards. This risk level setting mechanism allows vehicles to distinguish which risk events are immediate threats, which are potential hazards, and which are informative warnings, thus helping vehicles respond and make decisions more efficiently.

[0018] According to another aspect of the embodiments of this application, a vehicle control device is provided. The device includes: an acquisition unit, configured to acquire visibility data of the vehicle's current environment via vehicle-to-everything (V2X) technology in response to a vehicle-to-everything (V2X) data indicating the presence of a risk event; a determination unit, configured to determine weight coefficients corresponding to multiple sensors of the vehicle based on the visibility data; a processing unit, configured to perform weighted fusion processing on the perception data of the multiple sensors based on the weight coefficients corresponding to the multiple sensors to obtain a fusion result, wherein the perception data is used to characterize the road state of the road currently being traveled by the vehicle; a verification unit, configured to verify the risk event based on the fusion result to obtain a verification result; a generation unit, configured to generate a control command based on a predicted collision time between the vehicle and the target road object in response to a verification result indicating the presence of a target road object with a collision risk to the vehicle in the risk event; and a control unit, configured to control the vehicle based on the control command.

[0019] According to another aspect of the embodiments of this application, a vehicle is also provided, including a processor; a memory for storing processor-executable instructions; wherein the processor is configured to execute instructions to implement the method described above.

[0020] According to another aspect of the embodiments of this application, an electronic device is also provided, including a processor; a memory for storing processor-executable instructions; wherein the processor is configured to execute instructions to implement the method described above.

[0021] According to another aspect of the embodiments of this application, a computer-readable storage medium is provided, wherein a computer program is stored in the computer-readable storage medium, and the computer program is configured to perform the above-described method when run by a processor.

[0022] According to another aspect of the embodiments of this application, a computer program product is provided, including a computer program that, when executed by a processor, implements the above-described method.

[0023] It should be noted that the general descriptions above and the detailed descriptions that follow are merely examples and explanations of this application and do not constitute a limitation on this application. Attached Figure Description

[0024] Figure 1 This is a flowchart of a vehicle control method according to an embodiment of this application;

[0025] Figure 2 This is a flowchart of a vehicle control method based on V2X communication and multi-sensor fusion according to an embodiment of this application;

[0026] Figure 3 This is a flowchart of a method for determining the weighting coefficient of a sensor according to an embodiment of this application;

[0027] Figure 4 This is a flowchart of a distributed risk decision-making process according to an embodiment of this application;

[0028] Figure 5 This is a flowchart of another vehicle control method based on V2X communication and multi-sensor fusion according to an embodiment of this application;

[0029] Figure 6 This is a schematic diagram of a vehicle control device according to an embodiment of this application;

[0030] Figure 7 This is a structural diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0031] To make the technical problems, technical solutions, and beneficial effects solved by this application clearer, the following detailed description is provided in conjunction with embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0032] This application provides a vehicle control method. Please refer to the following embodiments. Figure 1 This includes the following steps:

[0033] Step S101: In response to the vehicle's Internet of Vehicles data indicating a risk event, obtain visibility data of the vehicle's current environment through Internet of Vehicles technology.

[0034] In the technical solution provided in step S101 of this application, the aforementioned vehicle network data includes communication data between vehicles and other vehicles, between vehicles and roadside equipment, and between vehicles and the cloud-based vehicle-to-everything (V2X) network. For example, the vehicle-to-everything (V2I) data may include communication data between vehicles and the infrastructure on the road they travel on (V2I), such as communication data between vehicles and roadside units (RSUs), traffic lights, weather stations, and other infrastructure. The V2I data may also include communication data between vehicles (V2V), such as communication data between vehicles and other vehicles on the road they travel on, used to share real-time location, speed, direction, and perceived environmental information. Furthermore, the V2I data may include communication data between vehicles and remote servers or the cloud (V2C), such as vehicles communicating with remote servers or the cloud to obtain broader data such as traffic conditions, road construction information, and weather forecasts. This is merely an example and does not limit the scope of the V2I data.

[0035] In this embodiment, vehicle-to-everything (V2X) data can be acquired through a V2X communication architecture. This V2X data can characterize whether a risk event exists for the vehicle. Such risk events can be dynamic traffic scenarios that pose a direct threat to the vehicle's driving safety, characterized by high urgency, high consequences, and quantifiable risk levels. When a risk event occurs, visibility data of the vehicle's current environment can be further obtained using V2X technology.

[0036] Optionally, the aforementioned vehicle-to-everything (V2X) data may include vehicle status information, vehicle environmental perception information, vehicle traffic information, and vehicle meteorological information. Vehicle status information may include vehicle position, speed, acceleration, and steering information. Environmental perception information may include the position and status of detected obstacles, pedestrians, and other vehicles. Traffic information includes traffic light status, road construction information, and traffic congestion. Meteorological information may include real-time weather conditions such as temperature, humidity, and visibility. The aforementioned risk events may include collision events, road construction events, and traffic light status events, with collision events having a higher risk level than road construction events, and road construction events having a higher risk level than traffic light status events.

[0037] Optionally, depending on the urgency of the risk events in the vehicle-to-everything (V2X) data, different communication channels can be used to receive the V2X data. For example, V2X data corresponding to high-risk events can be transmitted through a dedicated communication channel to ensure low latency and high priority; while V2X data corresponding to low-risk events can be received through a shared communication channel to balance resource usage and information flow.

[0038] Optionally, after identifying a risk event involving a vehicle, visibility data of the vehicle's current environment can be obtained further through vehicle-to-everything (V2X) technology. For example, by using communication technology between the vehicle and a roadside unit deployed on the road where the vehicle is currently traveling, the visibility data of the vehicle's current environment, collected in real time by the roadside unit, can be obtained.

[0039] For example, a high-precision weather sensor is deployed on the roadside unit along the vehicle's current route, continuously and stably monitoring visibility data in the area covered by the sensor. This visibility data characterizes the optical propagation conditions and visual perception limitations of the vehicle's current environment, and is a core indicator for measuring atmospheric transparency, light scattering, and occlusion intensity. It directly reflects the ability of visual and optical sensors (such as cameras and lidar) to clearly identify the distance and details of targets under adverse weather conditions such as rain, fog, snow, and dust storms. Lower visibility means more suspended particles or water vapor in the air, resulting in more severe light scattering or absorption, leading to blurred images, lost target outlines, sparse lidar point clouds, or even failure. Conversely, higher visibility indicates a clear environment and excellent sensor imaging quality. Therefore, visibility data quantifies the physical constraints of the environment on the actual working capabilities of the sensor perception system, providing an objective and quantifiable environmental basis for the vehicle to judge the reliability of the sensor's perception data.

[0040] In this step, by efficiently receiving and processing vehicle network data, if a risk event is identified, to avoid misjudgment, the reliability of the risk event can be verified by combining the perception data of multiple sensors of the vehicle. However, since the reliability of sensor perception data varies under different weather conditions, the visibility data of the current environment of the vehicle can be obtained through vehicle network technology, thereby providing reliable data basis for determining the reliability of perception data of multiple sensors in the subsequent risk event verification stage.

[0041] Step S102: Based on visibility data, determine the weight coefficients corresponding to the multiple sensors of the vehicle.

[0042] In the technical solution provided in step S102 of this application, as described above, visibility data is used to characterize the optical propagation conditions and visual perception limitations of the vehicle's current environment. This visibility data quantifies the physical constraints of the environment on the perception capabilities of multiple sensors in the vehicle. Based on this, the weighting coefficients corresponding to the vehicle's multiple sensors can be determined according to the visibility data of the vehicle's current environment. These weighting coefficients characterize the reliability of the perception data from the corresponding sensors in the vehicle's current environment.

[0043] In this embodiment, since different sensors perform differently under different visibility conditions, the weight coefficients of multiple sensors of the vehicle can be dynamically adjusted based on the visibility data of the current environment of the vehicle, so that the weight coefficients of multiple sensors are matched with the visibility of the current environment of the vehicle.

[0044] For example, in clear weather, cameras provide clear images and accurate target recognition information, thus receiving a higher weighting coefficient, while millimeter-wave radar and lidar receive lower weighting coefficients. In foggy or snowy weather, reduced visibility makes the camera's visual information blurry or even unreliable. However, millimeter-wave radar and lidar have strong penetration capabilities in adverse weather conditions and can maintain relatively stable performance. Therefore, in low-visibility situations, the weighting coefficients of these two types of sensors can be increased, while the weighting coefficient of the camera can be decreased.

[0045] Optionally, different visibility data correspond to different weather levels (e.g., from L1 to L5). Based on this, the weather level of the vehicle's current environment can be determined according to the visibility data of the vehicle's current environment. A weighted weather level mapping table is preset, which includes the mapping relationship between different weather levels and the weight coefficients corresponding to multiple sensors. Based on this, after determining the weather level of the vehicle's current environment, the weight coefficients corresponding to multiple sensors in the vehicle can be queried from the mapping table, thereby providing a data basis for subsequent data fusion of multiple sensors.

[0046] Optionally, during the data fusion stage, algorithms such as Kalman filtering can be used to weight the sensor data from each sensor according to the retrieved weight coefficients. This means that under adverse weather conditions, greater reliance will be placed on sensor data from millimeter-wave radar and lidar to generate more reliable and accurate environmental perception results, such as vehicle position, speed, and distance information.

[0047] In this step, by using the weighting coefficients of multiple sensors matched with the visibility of the vehicle's current environment, the perception data acquired by multiple sensors is fused to obtain more accurate fused perception data, thus providing a reliable data foundation for the subsequent verification of risk events.

[0048] Step S103: Based on the weight coefficients corresponding to the multiple sensors, the sensing data of the multiple sensors are weighted and fused to obtain the fusion result.

[0049] In the technical solution provided in step S103 of this application, after determining the weight coefficients corresponding to multiple sensors in the vehicle, the perception data of multiple sensors can be weighted and fused according to the weight coefficients to obtain a more accurate fusion result, thereby providing a reliable data foundation for the verification of subsequent risk events.

[0050] In this embodiment, multiple sensors are used to perceive the vehicle's surrounding environment from different dimensions. These multiple sensors include, but are not limited to: cameras (e.g., high-resolution forward-facing binocular cameras), millimeter-wave radar (e.g., 77GHz radar), lidar (e.g., 128-line lidar), and weather sensors. The cameras are used to acquire visual images of the road surface for target recognition and obstacle detection; the millimeter-wave radar is used to monitor the speed and distance information of objects around the vehicle, especially providing stable perception data under adverse weather conditions; the lidar is used to provide high-precision point cloud data of the vehicle's surrounding environment, helping to build a detailed road environment model; and the weather sensors are used to monitor environmental parameters such as temperature, humidity, and visibility in real time, and to adjust the weights in the sensor fusion algorithm.

[0051] Optionally, by acquiring perception data from multiple sensors, a more comprehensive and accurate road environment model can be constructed. Under normal driving conditions, cameras can provide clear visual information, but their effectiveness is greatly reduced in adverse weather conditions. In such cases, millimeter-wave radar and lidar, due to their low sensitivity to weather conditions, become the primary sources of perception data. This multi-sensor fusion approach can compensate for the limitations of a single sensor in specific environments, improving overall environmental perception capabilities.

[0052] Optionally, by using the determined weight coefficients corresponding to each sensor, weighted fusion processing can be performed on the perceived data from multiple sensors to obtain a more accurate perception data fusion result. For example, algorithms such as Kalman filtering can be used to weight the perceived data from each sensor according to the determined weight coefficients. This means that under adverse weather conditions, greater reliance is placed on the perception data from millimeter-wave radar and lidar to generate more reliable and accurate environmental perception results, such as information on vehicle position, speed, and distance.

[0053] In this step, by using the weighting coefficients of multiple sensors matched with the visibility of the vehicle's current environment, the perception data acquired by multiple sensors is fused to obtain more accurate fused perception data, thus providing a reliable data foundation for the subsequent verification of risk events.

[0054] Step S104: Based on the fusion results, verify the risk events to obtain the verification results.

[0055] In the technical solution provided in step S104 of this application, after obtaining the fusion result of the perception data of multiple sensors, the risk event can be verified based on the fusion result to obtain the verification result. The verification result can be used to indicate whether there is a target road object in the risk event that poses a collision risk to the vehicle.

[0056] In this embodiment, the fusion results can be used to detect whether there are road objects ahead of the vehicle that pose a collision risk, so as to verify the risk event and obtain the verification result.

[0057] Optionally, if a road object posing a collision risk to the vehicle is detected ahead, the target road object is identified as a potential collision risk in the risk event. Simultaneously, the vehicle's onboard camera and / or lidar are used to directly detect the target road object to confirm its position, size, and speed.

[0058] In this step, the risk event is verified by fusing the perception data from multiple sensors in the vehicle. The verification result can effectively improve the accuracy of risk event judgment, thereby reducing the possibility of false alarms and missed alarms.

[0059] Step S105: In response to the verification result indicating that there is a target road object in the risk event that poses a collision risk to the vehicle, a control command is generated based on the predicted collision time between the vehicle and the target road object.

[0060] In the technical solution provided in step S105 of this application, when it is determined from the verification result of step S104 that there is a target road object in the risk event that poses a collision risk to the vehicle, a control command can be generated based on the predicted collision time between the vehicle and the target road object.

[0061] In this embodiment, the time to collision (TTC) between the vehicle and the target road object can be predicted based on the position and speed information of the target road object, as well as the position and speed information of the vehicle. The predicted collision time is then used to generate control commands to control the vehicle.

[0062] Optionally, if the predicted collision time is less than a preset duration (e.g., 2 seconds), a first control command can be generated, such as activating the Automatic Emergency Braking (AEB) system to avoid or mitigate the collision. If the predicted collision time is greater than or equal to the preset duration, a second control command can be generated, such as controlling the vehicle to activate an audible and visual alarm.

[0063] In this step, the predicted collision time is introduced as a condition for the generation of control commands, enabling the system to decide whether to take emergency braking and other measures based on the real-time collision risk assessment results, thus avoiding the misjudgment that may be caused by the decision-making mechanism based on static thresholds.

[0064] Step S106: Control the vehicle based on the control command.

[0065] In the technical solution provided by step S106 of this application, after obtaining the control command through step S105, the vehicle can be controlled according to the control command.

[0066] In this embodiment, control commands can be sent to vehicle components via a Controller Area Network (CAN) bus to execute specific control actions and control the vehicle. The vehicle's actuators include, but are not limited to, Electronic Stability Control (ESC), Electric Power Steering (EPS), and Engine Management System.

[0067] Optionally, to prevent attacks on the CAN bus, encryption technology can be used to encrypt the control commands, ensuring that only legitimate vehicle components can decrypt and execute these commands, thus increasing the vehicle's security protection level.

[0068] Optionally, after the vehicle components execute control commands, the execution results can be collected, including braking distance, the actual effect of obstacle avoidance paths, etc., to evaluate the effectiveness of the control commands and the system's performance. Based on the feedback data of the execution results, the decision-making model can be updated and optimized online through federated learning technology, gradually reducing the misjudgment rate and improving the accuracy and efficiency of decision-making. This continuous learning and adaptation process enables the system to gradually improve its performance over time and achieve autonomous evolution.

[0069] In this step, a complete closed-loop mechanism is formed by controlling the execution of instructions, the feedback of execution results, and the online updating of the model, which ensures the accuracy and adaptability of the system's decision-making.

[0070] In steps S101 to S106 of this application, after determining that a risk event exists for the vehicle through vehicle-to-everything (V2X) data, visibility data of the vehicle's current environment can be further obtained through V2X technology. Based on this visibility data, the weighting coefficients of multiple sensors in the vehicle are dynamically adjusted to ensure the accuracy of the fusion result obtained when the perception data from multiple sensors is weighted and fused according to the weighting coefficients. Then, the risk event is further verified using the fusion result. Further verification of the risk event can determine whether there is a target road object in the risk event that poses a collision risk to the vehicle. If so, a control command is generated based on the predicted collision time between the vehicle and the target road object; based on the control command, the vehicle is controlled. In other words, in this embodiment of the application, dual detection is performed by fusing vehicle network data and perception data from multiple sensors of the vehicle. This can improve the detection accuracy of target road objects that pose a collision risk to the vehicle, reduce false alarms and false negatives, and then generate control commands based on the predicted collision time between the vehicle and the target road object to control the vehicle. This can improve the vehicle's driving safety and solve the technical problem in related technologies where the inability to accurately identify road objects that pose a collision risk to the vehicle in severe weather conditions leads to low vehicle driving safety.

[0071] The method described in this embodiment will now be further explained.

[0072] As an optional implementation, step S102, based on visibility data, determines the weight coefficients corresponding to multiple sensors of the vehicle, including: determining the weather level of the current environment of the vehicle based on visibility data; and determining the weight coefficients corresponding to multiple sensors from a weighted weather level mapping table based on the weather level, wherein the weighted weather level mapping table includes mapping relationships between various weather levels and sensor weight coefficients.

[0073] In this embodiment, the weighting coefficients corresponding to the aforementioned multiple sensors can be determined based on real-time acquired meteorological data. For example, after acquiring the visibility data of the vehicle's current environment, the meteorological level of the vehicle's current environment can be determined based on the visibility data. Then, according to a pre-established weighted meteorological level mapping table, the weighting coefficients corresponding to the multiple sensors in the current driving scenario can be determined to clarify the reliability of the perception data of the multiple sensors in the vehicle in the current environment.

[0074] Optionally, after obtaining the visibility data of the vehicle's current environment, the visibility data can be converted into a weather rating for that environment. For example, if the visibility data indicates that the vehicle's current environment has a clear environment with visibility greater than 200 meters, then the weather rating for that environment is determined to be L1. If the visibility data indicates that the vehicle's current environment has an extremely poor environment with visibility less than 20 meters, then the weather rating for that environment is determined to be L5. This is merely an example.

[0075] Optionally, after determining the weather level of the vehicle's current environment, a pre-set "weight-weather level mapping table" in the onboard unit can be consulted to determine the weight coefficients of multiple sensors matching that weather level. This weight-weather level mapping table is pre-constructed based on extensive measured data and sensor characteristic analysis, explicitly recording the weight coefficients of multiple sensors such as cameras, lidar, and millimeter-wave radar under each weather level, i.e., the perception reliability of those multiple sensors. For example, in a Level 1 clear weather environment, the camera provides clear images and is given a higher weight, while the radar, due to its relatively lower accuracy, has a lower weight. In Level 4 or Level 5 dense fog or blizzard environments, the camera is almost ineffective due to severely blurred images, and its weight is significantly reduced. The millimeter-wave radar, due to its strong penetration capability, is given the highest weight, and the lidar weight is correspondingly reduced but still higher than the camera. This mapping relationship is not a fixed average allocation but is dynamically configured based on the physical performance of the sensors under different weather conditions, ensuring that the fusion algorithm prioritizes the most reliable perception source in any environment. This mechanism enables adaptive perception and decision-making, making multi-sensor fusion no longer a simple data overlay, but an intelligent weighting based on environmental cognition, thereby significantly improving the accuracy and robustness of perception results.

[0076] Optionally, after obtaining the weight coefficients corresponding to multiple sensors, a fusion algorithm such as Kalman filtering can be used to weight the perception data acquired by multiple sensors to obtain the fusion result. This means that the perception data acquired by multiple sensors is no longer simply averaged, but weighted according to their reliability, giving more "trust" to those sensors that perform better in the current driving environment.

[0077] In this step, by adjusting the visibility data of the vehicle's current environment, the weighting coefficients of multiple sensors in the vehicle are dynamically adjusted. This allows for environmental perception based on the most reliable sensor data under various weather conditions, thereby improving the accuracy of perception. For example, in adverse weather conditions, such as fog, the system will rely more on data from millimeter-wave radar and lidar, as the performance of these sensors is not affected by reduced visibility and can provide stable target detection information. Based on the weighting coefficients of each sensor, the perceived data from multiple sensors is weighted and fused to obtain a fusion result, which can improve the accuracy of the fusion result.

[0078] As an optional implementation, step S104 verifies the risk event based on the fusion result to obtain a verification result, including: based on the fusion result, identifying road objects located in front of the vehicle on the road currently being driven by the vehicle to obtain an identification result, wherein the identification result is used to characterize whether there is a road object in front of the vehicle and the distance between the road object and the vehicle; in response to the identification result indicating that there is a road object in front of the vehicle and the distance between the road object and the vehicle is less than a preset distance threshold, determining that the verification result indicates that there is a target road object in the risk event that poses a collision risk to the vehicle.

[0079] In this embodiment, based on the fusion results of multiple sensors, road objects ahead of the vehicle on its current driving path are identified, yielding an identification result. Based on this identification result, potential road objects ahead of the vehicle can be effectively detected, and the distance between these road objects and the vehicle can be assessed.

[0080] Optionally, if the recognition result indicates that there is a road object in front of the vehicle and the distance between the road object and the vehicle is less than a preset distance threshold, then the verification result indicates that there is a target road object in the risk event that poses a collision risk to the vehicle.

[0081] Optionally, a preset distance threshold is used to determine whether a road object poses a collision risk. This preset distance threshold is set considering factors such as vehicle speed, road conditions, and the vehicle's braking performance to ensure timely identification and response to potential collision hazards under different circumstances.

[0082] Optionally, when the identification result indicates that there is a road object in front of the vehicle and the distance between the road object and the vehicle is less than a preset threshold, it can be determined that the vehicle is at risk of collision. That is, it can be determined that there is a target road object in the risk event that poses a collision risk to the vehicle.

[0083] In this step, road objects that pose a collision risk to vehicles are detected based on the fusion results of multiple sensors. This process can accurately identify road objects that pose a collision risk to vehicles and then verify the risk events.

[0084] As an optional implementation, a control command is generated based on the collision time between the vehicle and the target road object, including: determining the predicted collision time between the vehicle and the target road object based on the distance between them; generating a first control command in response to the predicted collision time being less than a preset duration, wherein the first control command is used to control the vehicle to brake suddenly; and generating a second control command in response to the predicted collision time being greater than or equal to the preset duration, wherein the second control command is used to control the vehicle to issue a warning.

[0085] In this embodiment, when generating control commands to control the vehicle based on the collision time between the vehicle and the target road object, the collision time between the vehicle and the target road object can be predicted first based on the distance between the vehicle and the target road object, and this collision time can be determined as the predicted collision time between the vehicle and the target road object. Then, the vehicle control commands can be generated based on the predicted collision time.

[0086] Optionally, a corresponding control command can be generated based on the comparison between the predicted collision time and the preset duration. For example, when the predicted collision time is less than the preset duration (e.g., 2 seconds), it means that a collision is about to occur, and emergency braking measures need to be taken immediately to avoid or mitigate the consequences of the collision. In this case, a first control command can be generated, which can be used to activate the vehicle's automatic emergency braking (AEB) system. This first control command can be sent to the vehicle's automatic emergency braking system via the CAN bus to cause the vehicle to decelerate rapidly.

[0087] Optionally, if the predicted collision time is greater than or equal to a preset duration, it indicates that the current collision risk is still within a controllable range and immediate braking is not required. However, in order to alert the driver of the vehicle to potential dangers, a second control command can be generated. This second control command can issue a warning signal to the driver through the vehicle's warning system (e.g., audible alarm, visual warning), prompting the driver to take appropriate driving behavior adjustments, such as slowing down or preparing to avoid an obstacle.

[0088] In this step, vehicle control commands are generated based on the predicted collision time between the vehicle and the target road object, enabling targeted vehicle control. If the predicted collision time is less than a preset duration, an emergency braking command is generated and executed, reducing the delay from risk identification to action and improving driving safety. In non-emergency situations with potential collision risks, the warning system alerts the driver, avoiding unnecessary emergency braking and promoting driver attention and response, thus contributing to improved overall traffic safety and smoothness. In other words, by dynamically deciding whether to apply emergency braking or only issue a warning based on the predicted collision time and a preset duration, the system not only improves decision-making accuracy but also demonstrates its intelligent adaptability to driving scenarios.

[0089] As an optional implementation, in step S101, before the vehicle network data indicates that a risk event has occurred in the vehicle, the method further includes: receiving vehicle network data through a dedicated communication channel in response to the risk level of the risk event being higher than a preset risk level threshold, wherein the dedicated communication channel is used to transmit communication data whose correlation with safe driving of the vehicle is higher than the preset threshold; and receiving vehicle network data through a shared communication channel in response to the risk level of the risk event being lower than or equal to the preset risk level threshold, wherein the shared communication channel is used to transmit communication data whose correlation with safe driving of the vehicle is lower than or equal to the preset threshold.

[0090] In this embodiment, vehicle network data can be transmitted through different communication channels according to the risk level of the risk events carried in the vehicle network data, thereby ensuring data transmission efficiency while maintaining the security and priority of data transmission.

[0091] Optionally, if the risk level of a risk event in the vehicle-to-everything (V2X) data exceeds a preset risk level threshold, it indicates a high risk level. In this case, the V2X information can be transmitted through a dedicated communication channel, allowing the vehicle to receive the V2X data. This dedicated communication channel is characterized by exclusive resource allocation, high priority, and low latency. It is specifically used to transmit communication data whose relevance to vehicle safety exceeds a preset threshold, such as emergency avoidance commands or high-risk warnings. This channel selection ensures that critical information can be rapidly and unimpededly transmitted to the vehicle system in emergency situations, enabling the vehicle to respond immediately.

[0092] Optionally, if the risk level of a risk event in the vehicle-to-everything (V2X) data is lower than or equal to a preset risk level threshold, it indicates that the risk level of the event is low. In this case, the V2X data will be transmitted through a shared communication channel. This shared communication channel allows more types of communication data to share limited network resources, such as traffic information updates and vehicle status reports. While this data is equally important, the requirements for immediate safety response are not as stringent as for high-risk events, therefore a slightly longer data transmission time is acceptable.

[0093] In this step, by differentiating communication channels for events of varying risk levels, not only is communication efficiency optimized, but safety is also enhanced. In emergency situations, rapid transmission of information for high-risk events ensures vehicles can take timely evasive action, while for low-risk events, although data transmission may be slightly delayed, network resources can still be effectively utilized through shared channels, avoiding waste of channel resources. Rationally allocating communication resources based on the risk level of events ensures the accurate transmission of information at critical moments.

[0094] As an optional implementation method, risk events are associated with risk levels.

[0095] In this embodiment, the risk level setting of a risk event directly affects the response speed, resource allocation, and overall system decision-making process. The risk level considers not only the urgency of the risk event but also its impact on driving safety and whether immediate action is required. By setting different risk levels for risk events, multi-source information can be processed more efficiently and rationally, ensuring timely and appropriate responses in complex traffic environments.

[0096] Optionally, risk levels are typically defined based on the severity and urgency of the event. The system assigns different risk levels to different risk events, allowing for priority handling of high-risk events when resources are limited.

[0097] Optionally, if the risk event is a "road ahead warning", its risk level can be marked as 0, i.e., the highest risk level, requiring a latency of ≤5ms. This means that the impact of this type of risk event on driving safety is the most direct and urgent, requiring immediate response and handling. A dedicated channel can be preempted to transmit this risk event. If the risk event is a "road construction notice", its risk level can be marked as 2, i.e., a relatively high risk level, requiring a latency of ≤100ms. Although this type of risk event also requires timely notification, its urgency and reaction time requirements are relatively low, and it can be transmitted through a shared channel contention access method. If the risk event is a "traffic light status notice", its risk level can be marked as 3, i.e., a low risk level, requiring a latency of ≤500ms. This type of information has a longer update cycle and a relatively smaller impact on immediate safety, so it can be transmitted through on-demand broadcasting.

[0098] Optionally, high-risk events (such as forward collision warnings) will be given the highest priority, thereby enabling them to preempt dedicated channels, achieve ultra-low latency transmission, and ensure that emergency braking commands can reach the target vehicle quickly.

[0099] Optionally, when handling multiple concurrent risk events, the order of priority can be determined based on the level of risk. For example, if both a forward collision warning and a road construction alert are received simultaneously, the forward collision warning can be processed first, as it has a higher risk level and a more direct impact on immediate safety.

[0100] Optionally, the risk level of a risk event also affects the activation of security protection mechanisms. For high-risk events, Trusted Execution Environment (TEE) hardware encryption and Simple Payment Verification (SPV) verification can be activated immediately to ensure the security and authenticity of control instructions and avoid interference from malicious information.

[0101] Optionally, by setting risk levels, it is possible to quickly identify which risk events require immediate response and which can be handled later, thereby improving the speed and efficiency of response in emergency situations.

[0102] The technical solutions of the embodiments of the present invention will be illustrated below with reference to preferred embodiments.

[0103] Figure 2 This is a flowchart of a vehicle control method based on V2X communication and multi-sensor fusion according to an embodiment of this application, as shown below. Figure 2 As shown, the method includes the following steps.

[0104] Step S201, Hardware deployment and environment calibration.

[0105] In this embodiment, multiple sensors are deployed at the front of the vehicle, including a forward-facing binocular camera, a 77GHz millimeter-wave radar, and a 128-line lidar. These sensors are responsible for capturing visual information, radar wave information, and 3D point cloud information around the vehicle, providing the vehicle with comprehensive environmental perception. Specifically, the binocular camera provides an RGB image stream, primarily used for target detection and tracking; the millimeter-wave radar detects the distance and speed of targets ahead by emitting and receiving electromagnetic waves; and the lidar creates a 3D point cloud map of the vehicle's surroundings using a high-speed rotating laser beam.

[0106] Optionally, roadside units (RSUs) integrating edge computing units and weather sensors can be deployed at road intersections or specific areas. The edge computing units (using NVIDIA Jetson AGX Orin as an example) are used for data processing and computing tasks, while the weather sensors are used to monitor environmental conditions in real time, such as temperature, humidity, and visibility, which are crucial for subsequent multi-sensor fusion processing.

[0107] Optionally, GPS-PPS signals can be used to synchronize the time of all devices, ensuring that the time deviation between the vehicle-side and roadside equipment is less than 1 microsecond. This is a prerequisite for multi-sensor data fusion, because the data must be in the same time coordinate system to be effectively weighted.

[0108] Optionally, a 5G NR-Uu interface was established between the OBU and RSU to ensure high-speed, low-latency communication connectivity. This interface can use fixed frequency bands and fixed bandwidths in the 5G frequency band, providing a channel for high-speed data exchange between vehicles and between vehicles and infrastructure, especially in emergency communications such as collision warnings, ensuring timely information delivery.

[0109] In this step, hardware deployment and environmental calibration provide a complete infrastructure for the collaborative decision-making system of intelligent connected vehicles, ensuring the accuracy of data collection and the efficiency of communication. This is a prerequisite for the normal operation of all subsequent perception, fusion, decision-making and execution functions.

[0110] Step S202: Multi-source data acquisition and preprocessing.

[0111] In this embodiment, multiple sensors on the vehicle can collect perception data. For example, RGB image streams are acquired through the vehicle's cameras, and target detection algorithms are used to identify vehicles, pedestrians, and obstacles. These are then combined with tracking algorithms for target tracking, providing structured visual information for subsequent fusion. The point cloud velocity matrix collected by millimeter-wave radar is processed through clustering and multi-target association algorithms to obtain stable target velocity and position information, compensating for the limitations of visual perception in adverse weather conditions. LiDAR receives point cloud data and uses feature extraction algorithms to extract the shape and position features of obstacles, enhancing the perception accuracy of nearby obstacles. Through image data preprocessing, visibility values ​​are quantified into discrete levels, providing clear environmental parameters for weight adjustment and ensuring that sensor data fusion can be dynamically optimized according to weather conditions.

[0112] Optionally, Table 1 is a mapping table between sensor type, data content, and processing method according to an embodiment of this application.

[0113] Table 1. Mapping Relationship between Sensor Type and Data Content / Processing Method

[0114]

[0115] Optionally, when preprocessing multi-source data, it is necessary to ensure that the raw data obtained from different sensors are converted into a format that the system can process uniformly, so as to facilitate the efficient execution of subsequent algorithms.

[0116] Optionally, preprocessing algorithms (such as YOLOv7, DBSCAN, PointPillars, etc.) can filter useful information from the raw data, remove noise, improve the accuracy and reliability of the data, and provide a solid foundation for subsequent decision-making.

[0117] Optionally, through preprocessing, the system can adjust the fusion strategy of sensor data according to real-time meteorological data, so that the sensing system can maintain high accuracy under different weather conditions such as rain, fog or sunny days.

[0118] Optionally, data preprocessing reduces the amount of data processed by subsequent fusion and decision-making algorithms, avoids the computation of redundant information, thereby improving the execution efficiency of the algorithm and shortening the decision response time.

[0119] Step S203: Multi-source data fusion.

[0120] In this embodiment, when performing fusion processing on multi-source sensor data collected by multiple sensors, the current weather conditions of the vehicle can be determined based on the visibility of the current environment. Then, different weight coefficients are assigned to the perception data of the vehicle's multiple sensors according to the weather conditions, and the multi-source data collected by multiple sensors is fused according to the weight coefficients to obtain the fusion result.

[0121] Optionally, Figure 3 This is a flowchart of a method for determining the weighting coefficients of a sensor according to an embodiment of this application, as shown below. Figure 3 As shown, the process begins with parameter definition, such as initializing meteorological parameters and sensor weighting coefficients, and checking for feasibility. Next, based on real-time collected meteorological data, the visibility of the vehicle's current environment is determined. If visibility < 50, the current weather condition is determined to be heavy fog; if 50 ≤ visibility ≤ 100, the current weather condition is determined to be light rain; and if visibility > 100, the current weather condition is determined to be clear. Then, based on the current weather conditions, weighting coefficients matching the current weather conditions are assigned to the vehicle's multiple sensors, and these weighting coefficients are returned. After determining the weighting coefficients for multiple sensors, the multi-source data collected by the multiple sensors can be fused based on these coefficients.

[0122] Alternatively, the Kalman filter algorithm can be used as a data fusion tool. Data from multiple sensors can be input into the filter, and the algorithm can estimate the state of the target, such as position, velocity, and orientation, based on the signal-to-noise ratio of each sensor and dynamically adjusted weights.

[0123] Step S204, V2X message dynamic scheduling.

[0124] In this embodiment, V2X messages are assigned different priorities based on the event type. For example, "Forward Collision Warning" is marked as the highest priority because this type of message is directly related to the safety of vehicles and people and needs to be communicated to the relevant vehicles in the shortest possible time. Other events, such as "Road Construction Notice" and "Traffic Light Status," are assigned lower priorities based on their impact on immediate safety.

[0125] Optionally, Table 2 is a message priority matrix table according to an embodiment of this application, which shows the mapping relationship between event type, risk level, latency requirement and protocol action.

[0126] Table 2 Message Priority Matrix

[0127]

[0128] Optionally, for high-priority messages, a "preemptive dedicated channel" approach is adopted. This means that after a high-risk event is detected, the vehicle's onboard unit (OBU) will immediately generate a message with a high-priority tag, such as Priority=0xE1. These messages will not enter the regular queuing process, but will directly preempt dedicated resource blocks in the communication network to achieve fast transmission and ensure end-to-end transmission of information within an extremely short time of ≤8ms.

[0129] In this step, by implementing a dynamic scheduling mechanism for V2X messages, the problem of excessive communication latency in traditional vehicle-to-everything (V2X) communication under congested scenarios is solved, improving the response speed and safety of intelligent connected vehicles in emergency situations. By rationally configuring communication resources and optimizing transmission processes, messages of different priorities can be effectively distinguished and processed.

[0130] Step S205, Distributed risk decision-making.

[0131] In this embodiment, distributed risk decision-making directly determines how the vehicle responds to potential emergencies to ensure driving safety. This step utilizes data collected and processed in previous steps, including environmental perception results from multi-sensor fusion, messages received from V2X communication, and the status of local safety protection mechanisms, to perform risk assessment and decision generation.

[0132] Optionally, Figure 4 This is a distributed risk decision-making flowchart according to an embodiment of this application, such as... Figure 4 As shown, after receiving a V2X alarm, the vehicle can use local sensors to verify the V2X warning information to determine if there is a target road object that poses a collision risk to the vehicle. If no such object exists, the verification fails, and the process ends. If the verification result indicates the presence of a target road object that poses a collision risk to the vehicle, the time to collision (TTC) between the target road object and the vehicle can be further assessed. If the TTC is less than 2 seconds, the Automatic Emergency Braking (AEB) system is triggered to perform emergency braking. If the TTC is greater than or equal to 2 seconds, an audible and visual alarm is triggered.

[0133] Step S206, Security Protection Linkage.

[0134] In this embodiment, the integrated security protection is a crucial element in ensuring communication security and preventing malicious attacks. This step focuses on how to achieve encrypted transmission and authenticity verification of commands both inside and outside the system through dual protection of hardware-level encryption and blockchain technology, thereby protecting the vehicle from cyberattacks and ensuring the safety of passengers and data.

[0135] Optionally, critical decision-making algorithms can be run within a secure enclave in the onboard unit (OBU) using a trusted execution environment (TEE). This ensures that the decision-making algorithms and the data they process remain secure even if the external software environment is attacked.

[0136] Optionally, encrypting control commands on the CAN bus using commercial cryptographic algorithms can effectively prevent unauthorized access and data tampering. This encryption layer is fundamental to the security of in-vehicle communications, protecting the command chain from the OBU to the vehicle execution layer (such as the Electronic Stability Control (ESC) and Electronic Power Steering (EPS)).

[0137] Optionally, the Roadside Unit (RSU) writes the hash value of the received V2X message into a pre-built consortium blockchain, which uses the Hyperledger Fabric platform. The participating nodes in the consortium blockchain include all registered OBUs and RSUs, which jointly maintain a distributed ledger to record the metadata of all high-priority messages.

[0138] Optionally, after receiving a message from the RSU or other vehicles, the On-Board Unit (OBU) uses a Simple Payment Verification (SPV) mechanism to verify the message's authenticity. SPV verification does not require downloading the entire blockchain; it only needs to check if the message's hash value exists in the consortium blockchain to quickly verify whether the message has been tampered with, thus improving the system's response speed.

[0139] Optionally, by combining hardware-level encryption technology and blockchain verification mechanisms, a complete and efficient intelligent connected vehicle security protection system has been constructed, which effectively resists possible cyberattacks, protects the vehicle's communication security, and is an indispensable key component of the entire collaborative decision-making system.

[0140] Step S207: Control command execution and closed-loop optimization.

[0141] In this embodiment, when the distributed risk decision-making module determines that a high-risk event exists (e.g., a collision warning), AEB will be triggered immediately, and the Electronic Stability Control (ESC) system will perform emergency braking to avoid or mitigate a collision as much as possible. At the same time, the Electric Power Steering (EPS) system may be activated, performing steering actions based on the obstacle avoidance path provided by the decision-making module to assist the vehicle in avoiding obstacles or dangerous areas.

[0142] Optionally, each action in the execution layer generates detailed feedback data, including the timestamp of the control command execution, vehicle ID, specific action performed (such as whether AEB was triggered), and execution result (such as the effectiveness of collision avoidance, deceleration, etc.), which is uploaded to the cloud or edge server in a standardized data format. This feedback data not only includes an evaluation of the effectiveness of the control command but also records multi-dimensional information such as changes in environmental conditions and the quality of sensor data, providing a data source for subsequent system optimization.

[0143] Optionally, a federated learning mechanism can be used to upload the collected success and failure cases to an edge server, where the server can centrally perform data analysis and model training to achieve regular updates to the decision-making model.

[0144] Through steps S201 to S207 above, the collaborative decision-making system based on V2X communication and multi-sensor fusion aims to solve the key challenges of intelligent connected vehicles in the perception, communication, and decision-making process, especially to enhance perception capabilities under adverse weather conditions, reduce communication latency in high-risk scenarios, improve system security, and achieve autonomous optimization of algorithms.

[0145] Figure 5 This is a flowchart of another vehicle control method based on V2X communication and multi-sensor fusion according to an embodiment of this application. Figure 5 As shown, after the vehicle system starts, it can first collect multi-source data through multiple sensors on the vehicle. Then, it analyzes the meteorological data in the multi-source data to determine the visibility of the current environment in which the vehicle is located. If the visibility is <50m, the radar weight is determined to be 0.8 and the camera weight to be 0.05. If the visibility is ≥100m, the radar weight is determined to be 0.7 and the camera weight to be 0.1.

[0146] After determining the weight coefficients corresponding to the sensors, the target list can be dynamically fused. Then, upon receiving the V2X message, its transmission priority can be determined based on the risk level of the event carried in the message. If the priority is 0, the V2X message has a high transmission priority; in this case, a dedicated communication channel can be preempted for transmission. If the priority is ≥2, the V2X message has a low transmission priority; in this case, the message can compete for access to a shared channel for transmission. Upon receiving the V2X message, the vehicle verifies it. If verification is successful, the collision time (TTC) between the target road object and the vehicle is determined. If TTC < 2s, AEB braking is triggered; if TTC ≥ 2s, an audible and visual alarm is triggered. Subsequently, the federated learning model is updated based on the vehicle's execution results.

[0147] This application also provides a vehicle control device 60, please refer to... Figure 6 The vehicle control device 60 includes: an acquisition unit 610, configured to acquire visibility data of the vehicle's current environment via vehicle-to-everything (V2X) technology in response to a vehicle network data indication of a risk event; a determination unit 620, configured to determine weight coefficients corresponding to multiple sensors of the vehicle based on the visibility data; a processing unit 630, configured to perform weighted fusion processing on the perception data of multiple sensors based on the weight coefficients corresponding to the multiple sensors to obtain a fusion result, wherein the perception data is used to characterize the road state of the road on which the vehicle is currently traveling; a verification unit 640, configured to verify the risk event based on the fusion result to obtain a verification result; a generation unit 650, configured to generate control commands based on the predicted collision time between the vehicle and the target road object in response to a verification result indicating that there is a target road object in the risk event that poses a collision risk to the vehicle; and a control unit 660, configured to control the vehicle based on the control commands.

[0148] Optionally, the determining unit 620 is further configured to: determine the weather level of the current environment of the vehicle based on visibility data; and determine the weight coefficients corresponding to multiple sensors from a weighted weather level mapping table based on the weather level, wherein the weighted weather level mapping table includes mapping relationships between various weather levels and sensor weight coefficients.

[0149] Optionally, the verification unit 640 is further configured to: based on the fusion result, identify road objects located in front of the vehicle on the road currently being driven by the vehicle, and obtain an identification result, wherein the identification result is used to characterize whether there are road objects in front of the vehicle and the distance between the road objects and the vehicle; in response to the identification result indicating that there are road objects in front of the vehicle and the distance between the road objects and the vehicle is less than a preset distance threshold, determine that the verification result indicates that there are target road objects in the risk event that pose a collision risk to the vehicle.

[0150] Optionally, the generation unit 650 is further configured to: determine the predicted collision time between the vehicle and the target road object based on the distance between the vehicle and the target road object; generate a first control command in response to the predicted collision time being less than a preset duration, wherein the first control command is used to control the vehicle to brake suddenly; and generate a second control command in response to the predicted collision time being greater than or equal to the preset duration, wherein the second control command is used to control the vehicle to issue a warning.

[0151] Optionally, the device 60 is further configured to: receive vehicle network data via a dedicated communication channel in response to a risk event whose risk level is higher than a preset risk level threshold, wherein the dedicated communication channel is used to transmit communication data whose correlation with safe vehicle driving is higher than the preset threshold; and receive vehicle network data via a shared communication channel in response to a risk event whose risk level is lower than or equal to the preset risk level threshold, wherein the shared communication channel is used to transmit communication data whose correlation with safe vehicle driving is lower than or equal to the preset threshold.

[0152] Optionally, risk events are associated with risk levels.

[0153] In the vehicle control device of this application, dual detection is performed by fusing vehicle network data and perception data from multiple sensors of the vehicle. This can improve the detection accuracy of target road objects that pose a collision risk to the vehicle, reduce false alarms and missed alarms, and then generate control commands based on the predicted collision time between the vehicle and the target road object to control the vehicle. This can improve the driving safety of the vehicle and solve the technical problem in related technologies where the inability to accurately identify road objects that pose a collision risk to the vehicle in severe weather conditions leads to low driving safety.

[0154] This application also provides an electronic device 70, please refer to... Figure 7 It includes a memory 710 and a processor 720, wherein the memory 710 is used to store computer programs; and the processor 720 is used to execute the programs stored in the memory 710 to implement the vehicle control method described in any embodiment of this application.

[0155] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the power distribution line protection method described in any embodiment of this application.

[0156] In this application, "multiple" refers to two or more.

[0157] In this application, unless otherwise expressly defined, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0158] The terms “first,” “second,” “third,” “fourth,” etc., in this application (if present) are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0159] In this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, in this application, the character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0160] Unless otherwise specified, all steps in this application may be performed sequentially or randomly. For example, if the method includes steps A and B, it means that the method may include steps A and B performed sequentially, or it may include steps B and A performed sequentially. For example, if the method may also include step C, it means that step C may be added to the method in any order. For example, the method may include steps A, B, and C, or it may include steps A, C, and B, or it may include steps C, A, and B, etc.

[0161] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for controlling a vehicle, characterized in that, include: In response to vehicle-to-everything (V2X) data indicating a risk event for the vehicle, visibility data of the vehicle's current environment is obtained through V2X technology. Based on the visibility data, the weighting coefficients corresponding to the multiple sensors of the vehicle are determined respectively; Based on the weight coefficients corresponding to the multiple sensors, the perception data of the multiple sensors are weighted and fused to obtain a fusion result, wherein the perception data is used to characterize the road condition of the road on which the vehicle is currently traveling; Based on the fusion results, the risk event is verified to obtain the verification results; In response to the verification result indicating the presence of a target road object in the risk event that poses a collision risk to the vehicle, a control command is generated based on the predicted collision time between the vehicle and the target road object; The vehicle is controlled based on the control commands.

2. The method according to claim 1, characterized in that, Based on the visibility data, the weighting coefficients corresponding to the multiple sensors of the vehicle are determined, including: Based on the visibility data, the weather level of the current environment in which the vehicle is located is determined; Based on the meteorological level, the weight coefficients corresponding to the multiple sensors are determined from the weighted meteorological level mapping relationship table, wherein the weighted meteorological level mapping relationship table includes the mapping relationship between multiple meteorological levels and sensor weight coefficients.

3. The method according to claim 1, characterized in that, Based on the fusion results, the risk event is verified to obtain verification results, including: Based on the fusion result, road objects located in front of the vehicle on the road currently being driven by the vehicle are identified to obtain an identification result, wherein the identification result is used to characterize whether the road object exists in front of the vehicle and the distance between the road object and the vehicle; In response to the recognition result indicating that there is a road object in front of the vehicle and the distance between the road object and the vehicle is less than a preset distance threshold, it is determined that the verification result indicates that there is a target road object in the risk event that poses a collision risk to the vehicle.

4. The method according to claim 3, characterized in that, Based on the predicted collision time between the vehicle and the target road object, control commands are generated, including: Based on the distance between the vehicle and the target road object, the predicted collision time between the vehicle and the target road object is determined; In response to the predicted collision time being less than a preset duration, a first control command is generated, wherein the first control command is used to control the emergency braking of the vehicle; In response to the predicted collision time being greater than or equal to the preset duration, a second control command is generated, wherein the second control command is used to control the vehicle to issue an alarm.

5. The method according to claim 1, characterized in that, The method further includes: In response to a risk event whose risk level is higher than a preset risk level threshold, the vehicle network data is received through a dedicated communication channel, wherein the dedicated communication channel is used to transmit communication data whose correlation with the safe driving of the vehicle is higher than a preset threshold; In response to the risk level of the risk event being lower than or equal to the preset risk level threshold, the vehicle network data is received through a shared communication channel, wherein the shared communication channel is used to transmit communication data whose correlation with the safe driving of the vehicle is lower than or equal to the preset threshold.

6. The method according to claim 5, characterized in that, Each risk event is associated with a risk level.

7. A vehicle control device, characterized in that, The device includes: The acquisition unit is used to acquire visibility data of the current environment of the vehicle in response to the vehicle's Internet of Vehicles data indicating that the vehicle has a risk event, through Internet of Vehicles technology. A determining unit is configured to determine the weighting coefficients corresponding to the multiple sensors of the vehicle based on the visibility data. The processing unit is used to perform weighted fusion processing on the perception data of the multiple sensors based on the weight coefficients corresponding to the multiple sensors respectively, and obtain a fusion result, wherein the perception data is used to characterize the road state of the road on which the vehicle is currently traveling; The verification unit is used to verify the risk event based on the fusion result and obtain the verification result; The generation unit is configured to, in response to the verification result indicating that there is a target road object in the risk event that poses a collision risk to the vehicle, generate control commands based on the predicted collision time between the vehicle and the target road object; A control unit is used to control the vehicle based on the control commands.

8. A vehicle, characterized in that, Including processor and memory, among which, Memory, used to store computer programs; A processor for executing a program stored in memory to implement the method described in any one of claims 1-6.

9. An electronic device, characterized in that, Including processor and memory, among which, Memory, used to store computer programs; A processor for executing a program stored in memory to implement the method described in any one of claims 1-6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1-6.

Images