Vehicle anti-collision method and device, electronic equipment and storage medium

Abstract

This application discloses a vehicle collision avoidance method, device, electronic device, and storage medium, comprising: acquiring target vehicle information and lane information behind the vehicle obtained based on heterogeneous sensors; acquiring vehicle information; identifying obstacle vehicles from the target vehicles behind the vehicle based on the vehicle information, target vehicle information, and lane information; determining the collision risk level between the vehicle and the obstacle vehicle based on the relative running state of the vehicle and the obstacle vehicle; and controlling the vehicle to execute a target collision avoidance strategy corresponding to the collision risk level according to a preset graded collision avoidance strategy. This application utilizes multiple heterogeneous sensors to achieve comprehensive and accurate perception of the environment behind the vehicle, quickly and accurately identify obstacle vehicles behind, conduct multi-dimensional risk assessment, and provide graded collision avoidance warnings and interventions based on the collision risk level, enabling the vehicle to actively and effectively avoid dangerous vehicles behind, thereby improving vehicle driving safety.

Description

Technical Field

[0001] This application belongs to the field of vehicle control technology, specifically relating to a vehicle collision avoidance method, device, electronic device, and storage medium. Background Technology

[0002] With the rapid development of automotive intelligent technology, Advanced Driver Assistance Systems (ADAS) have become an important feature of modern vehicles. Existing driver assistance functions such as Adaptive Cruise Control (ACC), Lane Keeping Assist (LKA), and Automatic Emergency Braking (AEB) primarily focus on environmental perception and hazard avoidance in front of and to the sides of the vehicle, with relatively limited ability to monitor and respond to vehicles approaching from behind. In complex road traffic environments, the failure of drivers to detect vehicles approaching at high speeds from behind due to limited visibility is one of the main causes of rear-end collisions, side impacts, and other traffic accidents.

[0003] Currently, most vehicle monitoring systems are based on a single type of sensor, such as relying solely on radar or solely on vision sensors. However, the accuracy of visual sensors in identifying dangerous targets behind vehicles drops significantly in adverse weather and lighting conditions, such as strong light, rain, snow, fog, or insufficient light. Radar is also insufficiently accurate in distinguishing between different types of targets, such as vehicles, guardrails, and non-motorized vehicles. Therefore, current monitoring systems based on a single type of sensor have limitations in perception in complex environments, making it difficult to accurately identify dangerous vehicles behind the vehicle. This prevents the vehicle from actively and effectively avoiding dangerous vehicles behind it, further affecting vehicle driving safety. Summary of the Invention

[0004] The purpose of this application is to provide a vehicle collision avoidance method, device, electronic device, and storage medium, which can solve the problem that current monitoring systems based on a single type of sensor have limited perception in complex environments, making it difficult to accurately identify dangerous vehicles behind the vehicle, resulting in the vehicle's inability to actively and effectively avoid dangerous vehicles behind it, and further affecting the vehicle's driving safety.

[0005] To solve the above-mentioned technical problems, this application is implemented as follows: In a first aspect, embodiments of this application provide a vehicle collision avoidance method, the method comprising: Acquire target vehicle information and lane information behind the vehicle based on heterogeneous sensors; the heterogeneous sensors include a visual camera, millimeter-wave radar, and lidar. Acquire vehicle information, and based on the vehicle information, the target vehicle information, and the lane information, identify the obstacle vehicle from the target vehicles behind the vehicle. The collision risk level between the vehicle and the obstacle vehicle is determined based on their relative operating states. According to the preset graded collision avoidance strategy, the vehicle is controlled to execute the target collision avoidance strategy corresponding to the collision risk level, wherein the graded collision avoidance strategy includes early warning prompts and / or avoidance responses.

[0006] Optionally, acquiring target vehicle information and lane information behind the vehicle based on heterogeneous sensors includes: Acquire environmental data behind the vehicle collected by various heterogeneous sensors; Based on the accuracy information of each heterogeneous sensor in the current environment, the weighting coefficients of the same environmental data collected by the heterogeneous sensors are determined. Based on the weighting coefficients, the same environmental data collected by the heterogeneous sensors are weighted and fused to obtain target vehicle information and lane information behind the vehicle.

[0007] Optionally, the step of acquiring vehicle information and identifying obstacle vehicles from target vehicles behind the vehicle based on the vehicle information, the target vehicle information, and the lane information includes: Obtain vehicle information, and generate the running trajectories of the vehicle and the target vehicle within the historical period based on the vehicle information and the target vehicle information, respectively. Based on the lane information, extract the geometric features of the lane where the vehicle is currently located; Based on the historical trajectories of the autonomous vehicle and the target vehicle, as well as the geometric features of the lane currently occupied by the autonomous vehicle, obstacle vehicles are identified from the target vehicles behind the autonomous vehicle.

[0008] Optionally, determining the collision risk level between the vehicle and the obstacle vehicle based on their relative operating states includes: Based on the vehicle information and the target vehicle information corresponding to the obstacle vehicle, the relative speed, relative acceleration, and relative distance between the vehicle and the obstacle vehicle are calculated. The collision time between the vehicle and the obstacle vehicle is calculated based on the relative speed, relative acceleration, and relative distance. Based on the relative speed and the collision time, the relative operating state of the vehicle and the obstacle vehicle is determined, and the relative operating state is compared with the preset judgment conditions to determine the collision risk level between the vehicle and the obstacle vehicle.

[0009] Optionally, controlling the vehicle to execute a target collision avoidance strategy corresponding to the collision risk level according to a preset graded collision avoidance strategy includes: Obtain each collision risk level and its corresponding preset graded collision avoidance strategy; Based on the preset graded collision avoidance strategy, determine the target collision avoidance strategy corresponding to the collision risk level; Generate warning prompts and / or hazard avoidance response commands corresponding to the target collision avoidance strategy, and control the vehicle to execute the target collision avoidance strategy according to the warning prompts and / or hazard avoidance response commands.

[0010] Optionally, determining the target collision avoidance strategy corresponding to the collision risk level based on a preset graded collision avoidance strategy includes: If the collision risk level is the first level, then the target collision avoidance strategy corresponding to the first level includes visual warning prompts; If the collision risk level is a second level, which is higher than the first level, then the target collision avoidance strategy corresponding to the second level includes visual warning prompts and audible warning prompts. If the collision risk level is a third level, which is higher than the second level, then the target collision avoidance strategy corresponding to the third level includes visual warning prompts, audible warning prompts, steering wheel vibration prompts, and hazard avoidance reminders. If the collision risk level is level four, which is higher than level three, then the target collision avoidance strategy corresponding to level four includes visual warning prompts, audible warning prompts, steering wheel vibration prompts, and active avoidance.

[0011] Optionally, the method further includes: When the collision avoidance strategy of the target includes active avoidance, it is determined that the vehicle needs to perform an active lane change; The system detects that the vehicle is in an authorized state for active risk avoidance and that the current environment of the vehicle meets the conditions for lane changing, and determines the active lane changing parameters; the active lane changing parameters include the target lane, the lateral angular velocity of the vehicle, and the steering angle; The vehicle is controlled to perform an active lane change according to the aforementioned active lane change parameters.

[0012] Secondly, embodiments of this application provide a vehicle collision avoidance device, the device comprising: The information acquisition module is used to acquire target vehicle information and lane information behind the vehicle based on heterogeneous sensors; the heterogeneous sensors include a visual camera, millimeter-wave radar and lidar. An obstacle recognition module is used to acquire vehicle information and, based on the vehicle information, the target vehicle information, and the lane information, identify obstacle vehicles from the target vehicles behind the vehicle. The risk assessment module is used to determine the collision risk level between the vehicle and the obstacle vehicle based on the relative operating status of the vehicle and the obstacle vehicle. The control module is used to control the vehicle to execute a target collision avoidance strategy corresponding to the collision risk level according to a preset graded collision avoidance strategy, wherein the graded collision avoidance strategy includes early warning prompts and / or hazard avoidance responses.

[0013] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the vehicle collision avoidance method as described in the first aspect.

[0014] Fourthly, embodiments of this application provide a readable storage medium storing a program or instructions that, when executed by a processor, implement the steps of the vehicle collision avoidance method as described in the first aspect.

[0015] The vehicle collision avoidance method provided in this application acquires target vehicle information and lane information behind the vehicle based on heterogeneous sensors, including a visual camera, millimeter-wave radar, and lidar. It acquires vehicle information and, based on the vehicle information, target vehicle information, and lane information, identifies obstacle vehicles from the target vehicles behind the vehicle. Based on the relative operating states of the vehicle and the obstacle vehicles, it determines the collision risk level between the vehicle and the obstacle vehicles. According to a preset graded collision avoidance strategy, it controls the vehicle to execute a target collision avoidance strategy corresponding to the collision risk level. The graded collision avoidance strategy includes warning prompts and / or hazard avoidance responses. This application embodiment utilizes multiple heterogeneous sensors to achieve comprehensive and accurate perception of the environment behind the vehicle, overcoming the limitations of a single sensor in complex environments. By employing information from the vehicle itself, the target vehicle, and lane information, it can quickly and accurately identify obstacle vehicles behind the vehicle without significantly increasing costs. This provides a reliable data foundation for autonomous risk avoidance assisted driving. By comprehensively considering the relative operating states of the vehicle and obstacle vehicles, such as relative speed, relative distance, approach time, and road conditions, a multi-dimensional risk assessment is conducted. This enables precise collision risk level classification and warning timing judgment, and provides graded collision avoidance warnings and interventions based on the collision risk level. This allows the vehicle to actively and effectively avoid dangerous vehicles behind it, further improving vehicle driving safety.

[0016] The above description is merely an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0017] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1This is a flowchart illustrating the steps of a vehicle collision avoidance method provided in an embodiment of this application; Figure 2 This is a schematic diagram of the architecture of a vehicle collision avoidance method provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of a vehicle collision avoidance device provided in an embodiment of this application; Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0020] The vehicle collision avoidance method, device, electronic device, and storage medium provided in this application will be described in detail below with reference to the accompanying drawings and through specific embodiments and application scenarios.

[0021] Reference Figure 1 The flowchart illustrates the steps of a vehicle collision avoidance method provided in an embodiment of this application. The method may include: Step 101: Obtain target vehicle information and lane information behind the vehicle based on heterogeneous sensors; heterogeneous sensors include visual cameras, millimeter-wave radar and lidar.

[0022] In this embodiment, in order to comprehensively and accurately perceive the environment behind the vehicle, the vehicle utilizes integrated multiple types of sensors, i.e., heterogeneous sensors, for environmental perception. The rear refers to the area behind the vehicle that the driver cannot directly observe through the rearview mirror inside the vehicle or the left and right rearview mirrors outside the vehicle. Specifically, it includes the area directly behind the rear of the vehicle, the left rear area, and the right rear area, which are usually located outside the B-pillar or C-pillar of the vehicle and their rearward extension. The heterogeneous sensors include visual cameras, millimeter-wave radar, and lidar. These heterogeneous sensors each have different perception characteristics and detection ranges. By fusing the data collected by these heterogeneous sensors, the accuracy and reliability of the vehicle's perception of the rear environment can be improved.

[0023] It should be noted that the vehicle anti-collision and anti-spinning method provided in this embodiment can be executed by the vehicle's controller or control system, etc. This embodiment uses the vehicle's domain controller as the execution subject. A high-performance domain controller, MDC620pro, can be used, with an AI computing power of 400 TOPS. The domain controller includes a multi-core CPU, GPU, and neural network processing unit for real-time processing of data collected by heterogeneous sensors and running complex data fusion algorithms and decision control algorithms. (Refer to...) Figure 2 This diagram illustrates the architecture of a vehicle collision avoidance method provided in an embodiment of this application. The heterogeneous sensors include a vision camera, millimeter-wave radar, and lidar. Based on a domain controller, it integrates multiple software functional modules, including sensor data fusion (for receiving, parsing, and synchronizing timestamps of data from various sensors, providing a unified internal data representation format); multi-target tracking (using an adaptive Kalman filter algorithm to continuously track detected target vehicles, reducing errors caused by measurement noise); and interfaces for risk assessment and decision-making, path planning, and vehicle control. Risk assessment and decision-making assesses potential collision risks in real time and determines appropriate avoidance strategies. Path planning plans a safe drivable trajectory when lane changes are needed to avoid collisions. Vehicle control calculates specific steering, acceleration, and braking commands. Finally, the execution layer executes a graded collision avoidance strategy according to the collision risk level determined by the domain controller.

[0024] In this embodiment, the heterogeneous sensors include a visual camera, millimeter-wave radar, and lidar. The visual camera can collect visual information about the environment behind the vehicle, including the color, texture, and shape of targets. It also collects image data behind the vehicle, including visual information such as target vehicles, lane lines, and road shoulders. The millimeter-wave radar transmits and receives millimeter-wave signals to collect information such as the distance, speed, and angle of target vehicles behind the vehicle. The lidar generates high-precision point cloud data by emitting laser beams and receiving reflected signals, used to collect high-precision point cloud data of target vehicles behind the vehicle, as well as point cloud data of lane lines, road shoulders, etc. For example, the vehicle can deploy one rear-view long-range millimeter-wave radar with a detection range ≥150 meters; two side-view mid-range millimeter-wave radars with a detection range ≥70 meters; one rear-view high-definition camera with a wide angle of over 120°; two side-view cameras; and one rear-view long-range solid-state lidar with a detection range ≥180 meters and a reflectivity of 10%.

[0025] In this embodiment, the target vehicle refers to other vehicles in the surrounding environment that may affect the vehicle's driving. This embodiment focuses on the target vehicle behind the vehicle. The target vehicle information includes the target vehicle's position, speed, acceleration, and direction of travel in the vehicle's coordinate system. Lane information refers to the position and type of lane lines in the environment where the vehicle is located. By fusing environmental data collected by heterogeneous sensors, the target vehicle information and lane information behind the vehicle are obtained. Specifically, the environmental data behind the vehicle collected by each heterogeneous sensor is acquired. Based on the accuracy information of each heterogeneous sensor in the current environment, the weight coefficient of the same environmental data collected by the heterogeneous sensors is determined. Based on the weight coefficient, the same environmental data collected by the heterogeneous sensors is weighted and fused to obtain the target vehicle information and lane information behind the vehicle. The multi-source sensor data fusion is not described in detail in this embodiment.

[0026] Step 102: Obtain vehicle information, and based on the vehicle information, target vehicle information, and lane information, identify the obstacle vehicle from the target vehicles behind the vehicle.

[0027] In this embodiment, vehicle information is acquired, and based on the vehicle information and target vehicle information, the running trajectories of the vehicle and target vehicle within a historical period are generated respectively. This allows for the identification of obstacle vehicles from among the target vehicles behind the vehicle, based on the vehicle information, target vehicle information, and lane information. Specifically, vehicle information can be directly obtained from the vehicle's domain controller. Vehicle information includes vehicle speed, yaw rate, acceleration, position, and direction of travel. Yaw rate is the speed at which the vehicle rotates around its vertical axis. Vehicle position is the vehicle's location in the vehicle coordinate system, typically with the rear axle center as the origin. Target vehicle information includes the target vehicle's position in the vehicle coordinate system, speed, acceleration, and direction of travel. Lane information is crucial for determining whether the target vehicle is in the same or adjacent lane as the vehicle. Lane information includes lane line positions and lane centerlines. Based on the vehicle information, target vehicle information, and lane information, it is determined whether a collision with the target vehicle is possible, thus identifying obstacle vehicles from among the target vehicles behind the vehicle.

[0028] Step 103: Determine the collision risk level between the vehicle and the obstacle vehicle based on the relative operating status of the vehicle and the obstacle vehicle.

[0029] In this embodiment, the relative running state between the vehicle and the obstacle vehicle is obtained based on the vehicle information and the target vehicle information. The relative speed, relative acceleration, and relative distance between the vehicle and the obstacle vehicle are calculated based on the vehicle information and the target vehicle information. The collision time between the vehicle and the obstacle vehicle is calculated based on the relative speed, relative acceleration, and relative distance. Thus, the relative running state between the vehicle and the obstacle vehicle is determined based on the relative speed and the collision time. The relative running state is then compared with a pre-set judgment condition to determine the collision risk level between the vehicle and the obstacle vehicle.

[0030] It should be noted that relative speed refers to the speed difference between the vehicle and the obstacle vehicle, relative acceleration refers to the acceleration difference between the vehicle and the obstacle vehicle, and relative distance refers to the spatial distance between the vehicle and the obstacle vehicle. By calculating the time to collision (TTC) between the vehicle and the obstacle vehicle, it can be determined whether there is a risk of collision. The collision time is the estimated time when a collision will occur when the vehicle and the obstacle vehicle are moving at their current relative speed and relative acceleration. The judgment criteria are that the relative speed and collision time fall within a pre-defined range of low risk, medium risk, high risk, and emergency risk. By comparing the relative speed and collision time between the vehicle and the obstacle vehicle with the pre-defined range of judgment criteria, the collision risk level between the vehicle and the obstacle vehicle is determined.

[0031] Step 104: According to the preset graded collision avoidance strategy, control the vehicle to execute the target collision avoidance strategy corresponding to the collision risk level, wherein the graded collision avoidance strategy includes warning prompts and / or avoidance responses.

[0032] In this embodiment, to ensure that the vehicle can take appropriate collision avoidance measures according to different collision risk levels, a target collision avoidance strategy corresponding to each collision risk level is determined based on a predefined graded collision avoidance strategy. This generates a warning instruction and / or hazard avoidance response instruction corresponding to the target collision avoidance strategy. The vehicle is then controlled to execute the target collision avoidance strategy based on the warning instruction and / or hazard avoidance response instruction. Specifically, the warning instruction is used to issue a warning to the driver, reminding them to pay attention to obstructing vehicles. The graded collision avoidance strategy includes a warning instruction and a hazard avoidance response. The warning instruction, which issues a warning to the driver to remind them to pay attention to obstructing vehicles, can include visual warnings, audible warnings, and steering wheel vibration warnings. The hazard avoidance response controls the vehicle to execute specific collision avoidance measures, which can include hazard avoidance reminders and active hazard avoidance.

[0033] The vehicle collision avoidance method provided in this application acquires target vehicle information and lane information behind the vehicle based on heterogeneous sensors, including a visual camera, millimeter-wave radar, and lidar. It acquires vehicle information and, based on the vehicle information, target vehicle information, and lane information, identifies obstacle vehicles from the target vehicles behind the vehicle. Based on the relative operating states of the vehicle and the obstacle vehicles, it determines the collision risk level between the vehicle and the obstacle vehicles. According to a preset graded collision avoidance strategy, it controls the vehicle to execute a target collision avoidance strategy corresponding to the collision risk level. The graded collision avoidance strategy includes warning prompts and hazard avoidance responses. This application embodiment utilizes multiple heterogeneous sensors to achieve comprehensive and accurate perception of the environment behind the vehicle, overcoming the limitations of a single sensor in complex environments. By employing information from the vehicle itself, the target vehicle, and lane information, it can quickly and accurately identify obstructing vehicles behind without significantly increasing costs. This provides a reliable data foundation for autonomous risk avoidance assisted driving. By comprehensively considering the relative operating states of the vehicle and obstructing vehicles, such as relative speed, relative distance, approach time, and road conditions, a multi-dimensional risk assessment is conducted. This enables precise collision risk level classification and warning timing judgment, and provides graded collision avoidance warnings and interventions based on the collision risk level. This allows the vehicle to actively and effectively avoid dangerous vehicles behind, further improving vehicle driving safety.

[0034] In some embodiments of this application, step 101, obtaining target vehicle information and lane information behind the vehicle based on heterogeneous sensors, may include the following steps: Sub-step 1011: Obtain environmental data behind the vehicle collected by each heterogeneous sensor; Sub-step 1012: Based on the accuracy information of each heterogeneous sensor in the current environment, determine the weighting coefficient of the same environmental data collected by the heterogeneous sensors. Sub-step 1013: Based on the weighting coefficient, the same environmental data collected by heterogeneous sensors is weighted and fused to obtain the target vehicle information and lane information behind the vehicle.

[0035] In this embodiment, since the detection accuracy of different sensors may vary under different environmental conditions, the detection accuracy of visual cameras decreases at night or in adverse weather conditions (such as rain or fog), while millimeter-wave radar and lidar can still maintain high detection accuracy under these conditions, but cannot distinguish target types. Therefore, this embodiment dynamically adjusts the weight coefficients of each sensor based on the current environmental conditions and sensor performance to achieve optimal data fusion, enabling comprehensive and accurate perception of the environment behind the vehicle and overcoming the limitations of a single sensor in complex environments.

[0036] In specific implementation, environmental data behind the vehicle is acquired from various heterogeneous sensors. This environmental data includes image data from a visual camera, specifically visual information such as target vehicles, lane lines, and road shoulders; distance, speed, and angle information of target vehicles behind the vehicle acquired by millimeter-wave radar; and point cloud data of target vehicles, lane lines, and road shoulders acquired by lidar. In this embodiment, because different sensors have different sampling frequencies and response times, the data from each heterogeneous sensor needs to be synchronized to ensure that the environmental data acquired by each sensor is at the same moment. Furthermore, since the data acquired by each sensor is usually located in its own sensor coordinate system, each sensor is pre-converted to the vehicle coordinate system during calibration to facilitate unified data processing and fusion.

[0037] In this embodiment, the weighting coefficients of the same environmental data collected by the heterogeneous sensors are determined based on the accuracy information of each heterogeneous sensor in the current environment. The accuracy information of the visual sensor includes image sharpness, contrast, signal-to-noise ratio, etc. The image quality can be evaluated by image processing algorithms such as edge detection and noise analysis. If the image quality is poor, such as blurry image or high noise, the weight of the data collected by the visual sensor is reduced. The accuracy information of radar sensors, including millimeter-wave radar and lidar, includes echo intensity and target resolution. Radar performance can be evaluated through radar signal processing algorithms such as multi-target tracking and echo intensity analysis. If the echo intensity is weak or the target resolution is low, the weight of the data collected by the radar sensor is reduced. For example, based on the accuracy information of various heterogeneous sensors in the current environment, a first weighting coefficient corresponding to the data collected by the vision sensor and a second weighting coefficient corresponding to the data collected by the radar are determined. If the accuracy of the vision sensor in detecting obstacles is higher than that of the radar in detecting obstacles, then the first weighting coefficient corresponding to the data collected by the vision sensor is greater than the second weighting coefficient corresponding to the data collected by the radar; if the accuracy of the vision sensor in detecting obstacles is lower than that of the radar in detecting obstacles, then the first weighting coefficient corresponding to the data collected by the vision sensor is less than the second weighting coefficient corresponding to the data collected by the radar. The weighting coefficients can be dynamically adjusted based on preset rules or requirements. In some embodiments, the impact of the vehicle's current speed on the sensor's accuracy is also considered. For example, when driving at low speeds, the weighting coefficients of the data collected by the visual sensor can be appropriately increased because the target moves slower at low speeds, giving the visual sensor more time to detect the target. When driving at high speeds, the weighting coefficients of the data collected by the radar can be appropriately increased because the target moves faster at high speeds, resulting in insufficient frequency and resolution for the visual sensor to capture images, thus enhancing the radar's real-time performance.

[0038] In practical implementation, after acquiring environmental data from various heterogeneous sensors and determining the weighting coefficients of each sensor's data, this environmental data needs to be weighted and fused to obtain more accurate target vehicle and lane information. Therefore, based on the weighting coefficients, the same environmental data acquired by heterogeneous sensors is weighted and fused to obtain target vehicle information and lane information behind the vehicle. For example, if the same environmental data represents the position of the target vehicle, the first position information of the target vehicle in the visual sensor coordinate system, such as the target's pixel coordinates, size, and shape, is acquired. The second position information of the target vehicle in the radar coordinate system, such as the target's distance, speed, angle, and 3D point cloud coordinates, is also acquired. The second position information in the radar coordinate system is mapped to the visual sensor coordinate system to obtain the third position information for unified processing. Based on the determined first weighting coefficient corresponding to the visual sensor data and the second weighting coefficient corresponding to the radar data, the first and third position information mapped to the same coordinate system are weighted and fused to obtain the final position information of the target vehicle.

[0039] This application embodiment achieves comprehensive and accurate perception of the environment behind the vehicle by fusing data from multiple heterogeneous sensors, overcoming the limitations of a single sensor in complex environments. It adopts dynamic weight allocation based on sensor accuracy to dynamically adjust the weights in data fusion, ensuring the reliability of perception information under various environmental conditions.

[0040] In some embodiments of this application, step 102, obtaining vehicle information and identifying obstacle vehicles from target vehicles behind the vehicle based on the vehicle information, target vehicle information, and lane information, may include the following steps: Sub-step 1021: Obtain vehicle information and generate the running trajectories of the vehicle and the target vehicle within the historical period based on the vehicle information and the target vehicle information, respectively. Sub-step 1022: Extract the geometric features of the lane currently occupied by the vehicle based on the lane information; Sub-step 1023: Based on the running trajectories of the self-vehicle and the target vehicle in the historical period and the geometric features of the lane currently occupied by the self-vehicle, identify the obstacle vehicle from the target vehicles behind the self-vehicle.

[0041] In this embodiment, by analyzing the running trajectories of the vehicle and the target vehicle, it is possible to determine whether the target vehicle may collide with the vehicle, thereby identifying potential obstacle vehicles. First, vehicle information is acquired. Based on the vehicle information and the target vehicle information, the running trajectories of the vehicle and the target vehicle in the historical period are generated respectively. Furthermore, based on lane information, the geometric features of the lane currently occupied by the vehicle are extracted. Thus, based on the running trajectories of the vehicle and the target vehicle in the historical period and the geometric features of the lane currently occupied by the vehicle, obstacle vehicles are identified from the target vehicles behind the vehicle.

[0042] In the specific implementation, the vehicle information can be directly obtained from the vehicle's domain controller. This information includes vehicle speed, yaw rate, acceleration, position, and direction of travel. Yaw rate is the speed at which the vehicle rotates around its vertical axis, and the vehicle position is its location in the vehicle coordinate system, typically with the rear axle center as the origin. The target vehicle information includes its position, speed, acceleration, and direction of travel in the vehicle coordinate system. Based on the vehicle and target vehicle information, the trajectory of each vehicle within a historical period is generated. The vehicle's trajectory is calculated by connecting its historical positions from the previous N periods based on its speed, yaw rate, acceleration, and direction, reflecting its recent motion trend and path. The target vehicle's trajectory is calculated by connecting its historical positions from the previous N periods based on its speed, acceleration, and direction, reflecting its recent motion trend and path.

[0043] In this embodiment, lane information is a crucial basis for determining whether a target vehicle is in the same or adjacent lane as the vehicle. By extracting the geometric features of the lane, the relative positional relationship between the target vehicle and the vehicle can be accurately analyzed. Lane information includes lane line positions and the lane centerline, which is the midline between the two lane lines. Based on the lane information, the geometric features of the lane currently occupied by the vehicle are extracted. The distance between the two lane lines is calculated using the lane line positions to obtain the lane width. The lateral intercept is obtained based on the lateral distance between the vehicle's current position and the lane centerline. The lane width, lane centerline, and lateral intercept define the geometric features of the lane currently occupied by the vehicle, providing a benchmark for subsequent analysis of the relative positional relationship between the target vehicle and the vehicle.

[0044] In this embodiment, by analyzing the trajectories of the self-vehicle and the target vehicle within a historical period, as well as the geometric characteristics of the lane currently occupied by the self-vehicle, it is possible to determine whether the target vehicle is likely to collide with the self-vehicle in the future, thereby identifying the obstacle vehicle. The historical period can be set according to actual needs, such as a time window within the past 3 seconds. An obstacle vehicle refers to a stationary or moving vehicle that physically obstructs, poses a collision risk, or interferes with the normal driving path of the self-vehicle within the traffic space of the vehicle's driving road. For example, if the trajectory extension line of the target vehicle within the historical period intersects with the trajectory extension line of the self-vehicle within the same lane, the target vehicle may become an obstacle vehicle. Here, the trajectory extension line refers to the future short-term driving path predicted based on the target vehicle's speed and driving direction within the historical period, combined with its lane. If the trajectory of the target vehicle within the historical period is close to that of the self-vehicle within the historical period in an adjacent lane, the target vehicle may become a potential obstacle vehicle. In some embodiments, the speed, acceleration, and direction of the target vehicle can also be considered to determine the relative motion state between the target vehicle and the owner vehicle, thereby identifying the obstacle vehicle. If the target vehicle is accelerating towards the owner vehicle, the target vehicle may become an obstacle vehicle; if the target vehicle is decelerating away from the owner vehicle, the target vehicle may not be an obstacle vehicle.

[0045] The embodiments of this application use vehicle information, target vehicle information, and lane information, which can quickly and accurately identify vehicles with obstacles behind without significantly increasing costs, providing a reliable data foundation for autonomous risk avoidance assisted driving.

[0046] In some embodiments of this application, step 103, determining the collision risk level between the vehicle and the obstacle vehicle based on their relative operating states, includes: Sub-step 1031: Based on the vehicle information and the target vehicle information corresponding to the obstacle vehicle, calculate the relative speed, relative acceleration and relative distance between the vehicle and the obstacle vehicle. Sub-step 1032: Calculate the collision time between the vehicle and the obstacle vehicle based on the relative velocity, relative acceleration, and relative distance. Sub-step 1033: Based on relative speed and collision time, determine the relative running state of the vehicle and the obstacle vehicle, and compare the relative running state with the pre-set judgment conditions to determine the collision risk level between the vehicle and the obstacle vehicle.

[0047] In this embodiment, the relative running state between the vehicle and the obstacle vehicle is crucial for assessing collision risk. By calculating the relative speed, relative acceleration, and relative distance between the vehicle and the obstacle vehicle, basic data is provided for subsequent collision risk assessment. Therefore, based on the vehicle information and the target vehicle information corresponding to the obstacle vehicle, the relative speed, relative acceleration, and relative distance between the vehicle and the obstacle vehicle are calculated. Based on the relative speed, relative acceleration, and relative distance, the collision time between the vehicle and the obstacle vehicle is calculated. Thus, based on the relative speed and collision time, the relative running state between the vehicle and the obstacle vehicle is determined, and the relative running state is compared with pre-set judgment conditions to determine the collision risk level between the vehicle and the obstacle vehicle.

[0048] In practice, relative speed refers to the speed difference between the vehicle and the obstacle vehicle; relative acceleration refers to the acceleration difference between the vehicle and the obstacle vehicle; and relative distance refers to the spatial distance between the vehicle and the obstacle vehicle. By calculating the time to collision (TTC) between the vehicle and the obstacle vehicle, it can be determined whether there is a risk of collision. The collision time is the estimated time when a collision will occur when the vehicle and the obstacle vehicle are moving at their current relative speed and relative acceleration. Specifically, the collision time between the vehicle and the obstacle vehicle is calculated based on the relative speed, relative acceleration, and relative distance using the following formula:

[0049] In this embodiment, by analyzing the relative speed and collision time between the host vehicle and the obstacle vehicle, the relative running state between the host vehicle and the obstacle vehicle is obtained, and the relative running state is compared with the preset determination conditions to determine the collision risk level between the host vehicle and the obstacle vehicle. The determination of the collision risk level provides an important basis for the decision-making of vehicle collision avoidance. The determination conditions are the range of intervals set for relative speed and collision time in low risk, medium risk, high risk, and emergency risk. The relative speed and collision time between the host vehicle and the obstacle vehicle are compared with the range of intervals of the preset determination conditions to determine the collision risk level between the host vehicle and the obstacle vehicle. For example, referring to Table 1 below: The relative speed and collision time are respectively set with ranges of intervals in low risk, medium risk, high risk, and emergency risk. If TTC > 7s and relative speed < 20 km / h, the collision risk level between the host vehicle and the obstacle vehicle is level 1, i.e., low risk. If 5s < TTC ≤ 7s and 20 km / h ≤ relative speed < 30 km / h, the collision risk level between the host vehicle and the obstacle vehicle is level 2, i.e., medium risk. If 3s < TTC ≤ 5s and 30 km / h ≤ relative speed < 50 km / h, the collision risk level between the host vehicle and the obstacle vehicle is level 3, i.e., high risk. If TTC ≤ 3s and relative speed ≥ 50 km / h, the collision risk level between the host vehicle and the obstacle vehicle is level 4, i.e., emergency risk. Specifically, the thresholds and ranges of collision time and relative speed in the determination conditions can be adjusted according to the actual situation or requirements, and no specific limitation is made here.

[0050] Table 1: Corresponding relationship table between collision risk level and hierarchical collision avoidance strategy

[0051] The embodiment of the present application accurately evaluates the collision risk level between the host vehicle and the obstacle vehicle, can quickly and accurately judge the collision risk between the host vehicle and the obstacle vehicle, and the determination of the collision risk level provides an important basis for the decision-making of vehicle collision avoidance.

[0052] In some embodiments of the present application, in step 104, according to the preset hierarchical collision avoidance strategy, controlling the host vehicle to execute the target collision avoidance strategy corresponding to the collision risk level may include the following steps: Sub-step 1041, obtaining each collision risk level and the corresponding preset hierarchical collision avoidance strategy; Sub-step 1042, determining the target collision avoidance strategy corresponding to the collision risk level according to the preset hierarchical collision avoidance strategy; Sub-step 1043, generating a warning prompt instruction and / or a risk avoidance response instruction corresponding to the target collision avoidance strategy, and controlling the host vehicle to execute the target collision avoidance strategy according to the warning prompt instruction and / or the risk avoidance response instruction.

[0053] In this embodiment, there is a correspondence between collision risk levels and collision avoidance strategies. To ensure that the vehicle can take appropriate collision avoidance measures according to different collision risk levels, this embodiment predefines collision avoidance strategies corresponding to each collision risk level. The specific graded collision avoidance strategies corresponding to each collision risk level are shown in Table 1 above. The graded collision avoidance strategies include warning prompts and hazard avoidance responses. Specifically, each collision risk level and its corresponding preset graded collision avoidance strategy are obtained. Based on the preset graded collision avoidance strategy, a target collision avoidance strategy corresponding to the collision risk level is determined, thereby generating a warning prompt instruction and / or hazard avoidance response instruction corresponding to the target collision avoidance strategy. Based on the warning prompt instruction and / or hazard avoidance response instruction, the vehicle is controlled to execute the target collision avoidance strategy.

[0054] In this embodiment, based on the determined collision risk level, a corresponding target collision avoidance strategy is selected from a preset tiered collision avoidance strategy to address the current collision risk. The target collision avoidance strategy can take the most appropriate collision avoidance measures based on the current collision risk level to ensure the vehicle's safety. After the target collision avoidance strategy is determined, corresponding warning prompts and / or avoidance response commands are generated, and the vehicle is controlled to execute the target collision avoidance strategy through these warning prompts and / or avoidance response commands. Specifically, the warning prompts are used to issue warnings to the driver, reminding them to pay attention to obstructing vehicles; the warning prompts may include visual warnings, audible warnings, and steering wheel vibration prompts. The avoidance response commands are used to control the vehicle to execute specific collision avoidance measures; the avoidance response commands may include avoidance reminders, active avoidance, etc.

[0055] This application provides graded collision avoidance warnings and interventions based on collision risk levels, ensuring that the vehicle can take appropriate collision avoidance measures according to different collision risk levels, thereby enabling the vehicle to actively and effectively avoid dangerous vehicles behind it, and further improving vehicle driving safety.

[0056] In some embodiments of this application, sub-step 1042, determining a target collision avoidance strategy corresponding to the collision risk level based on a preset graded collision avoidance strategy, may specifically include: If the collision risk level is Level 1, the target collision avoidance strategy corresponding to Level 1 includes visual warning prompts; If the collision risk level is Level 2, which is higher than Level 1, then the target collision avoidance strategy corresponding to Level 2 includes visual warning prompts and audible warning prompts. If the collision risk level is Level 3, which is higher than Level 2, then the target collision avoidance strategies corresponding to Level 3 include visual warning prompts, audible warning prompts, steering wheel vibration prompts, and hazard avoidance reminders. If the collision risk level is Level 4, which is higher than Level 3, then the target collision avoidance strategies corresponding to Level 4 include visual warning prompts, audible warning prompts, steering wheel vibration prompts, and active avoidance.

[0057] In this embodiment of the application, to ensure that appropriate anti-collision measures are taken according to different collision risk levels, anti-collision strategies corresponding to each collision risk level are predefined to achieve dynamic response to collision risks. The collision risk levels are divided into four levels, from low to high: Level 1, Level 2, Level 3, and Level 4. The anti-collision strategies corresponding to each level are as follows: If the collision risk level is Level 1, which is low risk, it means that the risk of collision between the vehicle and an obstacle vehicle is low and no immediate action is required, but monitoring is still necessary. The target collision avoidance strategy corresponding to Level 1 includes visual warning prompts. Visual warning prompts are slight visual prompts issued by lights inside or outside the vehicle, such as dashboard warning lights or head-up displays. These can be green icons, reminding the driver to pay attention to obstacle vehicles behind. They only provide visual prompts and do not involve specific avoidance responses, keeping the driver alert.

[0058] If the collision risk level is Level 2 (Medium Risk), which is higher than Level 1, it indicates that there is a certain risk of collision between the vehicle and an obstacle vehicle. Close attention and early warning are required. The target collision avoidance strategy corresponding to Level 2 includes visual warning prompts and audible warning prompts. Visual warning prompts are more obvious visual prompts issued by lights inside or outside the vehicle, which can be yellow icons. Audible warning prompts can be alarm sounds issued by the vehicle's audio system to remind the driver to pay attention to obstacle vehicles behind. Combining visual and audible warnings enhances the driver's perception ability, but does not involve specific avoidance responses.

[0059] If the collision risk level is Level 3 (high risk), which is higher than Level 2, it indicates that there is a high risk of collision between the vehicle and an obstacle vehicle. Immediate warning measures are required, and driver intervention may be necessary. The target collision avoidance strategies corresponding to Level 3 include visual warning prompts, audible warning prompts, steering wheel vibration prompts, and hazard avoidance reminders. Visual warning prompts provide strong visual prompts through lights inside or outside the vehicle, which may be red icons. Audible warning prompts emit a continuous alarm sound through the vehicle's audio system. Steering wheel vibration prompts use vibrations from the steering wheel to alert the driver to obstacle vehicles behind. Hazard avoidance reminders can provide the driver with specific hazard avoidance operation suggestions through the instrument panel or head-up display, such as "Please slow down" or "Please change lanes." Combining visual, audible, vibration, and operational prompts, the system comprehensively alerts the driver and guides them to take hazard avoidance measures.

[0060] If the collision risk level is Level 4, which is higher than Level 3 and is considered an emergency risk, it indicates that there is an extremely high risk of collision between the vehicle and the obstacle vehicle, requiring immediate evasive action. The driver may not be able to react in time. The target collision avoidance strategy corresponding to Level 4 includes visual warning prompts, audible warning prompts, steering wheel vibration prompts, and active avoidance. Based on visual warning prompts, audible warning prompts, and steering wheel vibration prompts, active avoidance is implemented, which means automatically executing avoidance operations. For example, under safe conditions, the vehicle may quickly change lanes to avoid the obstacle vehicle. By combining visual, audible, vibration, and active avoidance operations, the vehicle can actively execute avoidance measures to prevent a collision if the driver cannot react in time.

[0061] This application embodiment uses a graded collision avoidance strategy response mechanism to ensure that the vehicle can adopt appropriate collision avoidance strategies under different collision risk levels. This avoids excessive interference with driver operation and can automatically perform evasive actions when necessary, balancing the driver's right to know with the need for rapid response in emergency situations, and further improving the effectiveness of vehicle collision avoidance.

[0062] In some embodiments of this application, the method may further include: Step 01: If the target collision avoidance strategy includes active avoidance, determine that the vehicle needs to perform an active lane change. Step 02: Detect that the vehicle is in an authorized state for active risk avoidance and that the current environment of the vehicle meets the conditions for lane changing, and determine the active lane changing parameters; the active lane changing parameters include the target lane, the lateral angular velocity of the vehicle, and the steering angle; Step 03: Control the vehicle to perform active lane change according to the active lane change parameters.

[0063] In this embodiment, when a rapidly approaching obstacle vehicle is detected from behind and poses an emergency risk, the vehicle needs to take active avoidance measures. Lane changing is one of the common avoidance strategies. Therefore, when the avoidance response of the target collision avoidance strategy includes active avoidance, it is determined that the vehicle needs to perform an active lane change. However, the lane change operation must meet certain safety conditions and ensure smoothness and safety during execution. Specifically, the vehicle is monitored to be in an authorized state for active avoidance, and the current environment of the vehicle meets the lane change conditions. The active lane change parameters are determined. The active lane change parameters include the target lane, the lateral angular velocity of the vehicle, and the steering angle, thereby controlling the vehicle to perform an active lane change according to the active lane change parameters.

[0064] In practice, the first step is to confirm that the current environment meets the conditions for lane changing. These conditions specifically include sufficient space in the target lane, the absence of other vehicles in the target lane, and road conditions allowing the lane change operation. In other words, the target lane must be spacious enough to allow the vehicle to complete the lane change without interfering with other vehicles in the target lane, and there must be no other vehicles in close proximity or a sufficient safety distance. This can be achieved by using the vehicle's sensors, such as cameras, millimeter-wave radar, or lidar, to detect the space in the target lane. Road conditions allowing the lane change operation include, for example, non-solid-line lanes, non-congested road sections, and non-merging intersections and other critical areas, as merging intersections and ramp entrances are generally unsuitable for lane changes. In this embodiment, if the collision avoidance strategy includes active avoidance, it is necessary to determine whether to adopt active lane changing as an avoidance strategy. Based on the collision risk level and the current environmental conditions, it is determined whether lane changing is the best avoidance strategy. If lane changing is the best avoidance strategy, it is determined that the vehicle will perform active lane changing. The system confirms that the vehicle is in an authorized state for active risk avoidance, meaning the system has the authority to perform an active lane change operation. It also monitors whether the vehicle's current environment meets the lane change conditions, including the space in the target lane, the vehicle situation within the target lane, and road conditions. This confirms the vehicle is in an authorized state for active risk avoidance and that the current environment meets the lane change conditions. The system then determines the active lane change parameters. By sensing the vehicle's surrounding environment in real time, and based on the direction of the target lane and the vehicle's current position, a lane change trajectory is generated. This trajectory needs to consider the vehicle's lateral angular velocity, steering angle, and dynamic constraints during the lane change process, such as maximum lateral acceleration and maximum steering angle. Therefore, the active lane change parameters include the target lane, the vehicle's lateral angular velocity, and the steering angle. The target lane is the lane the vehicle needs to change to. The vehicle's lateral angular velocity is calculated based on the direction of the target lane and the vehicle's current position. The steering angle is calculated based on the direction of the target lane and the vehicle's current position. These details will not be elaborated upon here.

[0065] In this embodiment, after determining the active lane change parameters, the vehicle is controlled to perform a lane change operation according to the active lane change parameters. The lateral movement of the vehicle is precisely controlled by vehicle control execution modules such as the steering system and braking system. The control of lateral movement includes steering control, that is, controlling the steering system of the vehicle to complete the lane change according to the steering angle parameters, and lateral acceleration control, that is, controlling the lateral acceleration of the vehicle according to the lateral angular velocity parameters to ensure a smooth lane change process. In addition, the dynamics of rapidly approaching vehicles behind and the situation of the target lane can be continuously monitored during the lane change process. If a new risk is detected, such as the appearance of a new vehicle in the target lane, the vehicle needs to adjust the lane change strategy or abort the lane change. During the active lane change process, the system allows the driver to intervene and terminate the operation at any time.

[0066] When the embodiment of this application detects that an obstacle vehicle is rapidly approaching from behind and poses an emergency risk, the vehicle takes active avoidance measures to effectively avoid the obstacle vehicle in an emergency. By precisely controlling the lateral movement of the vehicle, the lane change process is ensured to be smooth and safe, thereby improving vehicle safety.

[0067] Reference Figure 3 The diagram shows a structural schematic of a vehicle collision avoidance device according to an embodiment of this application. The device includes: The information acquisition module 201 is used to acquire target vehicle information and lane information behind the vehicle based on heterogeneous sensors; the heterogeneous sensors include a visual camera, millimeter-wave radar and lidar. The obstacle recognition module 202 is used to acquire vehicle information and, based on the vehicle information, the target vehicle information, and the lane information, identify the obstacle vehicle from the target vehicles behind the vehicle. Risk assessment module 203 is used to determine the collision risk level between the vehicle and the obstacle vehicle based on the relative operating status of the vehicle and the obstacle vehicle. The control module 204 is used to control the vehicle to execute a target collision avoidance strategy corresponding to the collision risk level according to a preset graded collision avoidance strategy, wherein the graded collision avoidance strategy includes early warning prompts and / or avoidance responses.

[0068] Optionally, the information acquisition module 201 includes: The first acquisition submodule is used to acquire environmental data behind the vehicle collected by various heterogeneous sensors; The first determining submodule is used to determine the weighting coefficient of the same environmental data collected by the heterogeneous sensors based on the accuracy information of each heterogeneous sensor in the current environment. The fusion submodule is used to perform weighted fusion of the same environmental data collected by the heterogeneous sensors based on the weight coefficients to obtain target vehicle information and lane information behind the vehicle.

[0069] Optionally, the obstacle recognition module 202 includes: A generation submodule is used to acquire vehicle information and generate the running trajectories of the vehicle and the target vehicle in the historical period based on the vehicle information and the target vehicle information, respectively. The extraction submodule is used to extract the geometric features of the lane where the vehicle is currently located based on the lane information. The identification submodule is used to identify obstacle vehicles from the target vehicles behind the vehicle based on the running trajectories of the vehicle and the target vehicle in the historical period and the geometric features of the lane where the vehicle is currently located.

[0070] Optionally, the risk assessment module 203 includes: The first calculation submodule is used to calculate the relative speed, relative acceleration, and relative distance between the vehicle and the obstacle vehicle based on the vehicle information and the target vehicle information corresponding to the obstacle vehicle. The second calculation submodule is used to calculate the collision time between the vehicle and the obstacle vehicle based on the relative speed, relative acceleration and relative distance. The second determining submodule is used to determine the relative running state of the vehicle and the obstacle vehicle based on the relative speed and the collision time, and compare the relative running state with the preset judgment conditions to determine the collision risk level between the vehicle and the obstacle vehicle.

[0071] Optionally, the control module 204 includes: The second acquisition submodule is used to acquire each collision risk level and the corresponding preset graded collision avoidance strategy; The third determining submodule is used to determine the target collision avoidance strategy corresponding to the collision risk level based on the preset hierarchical collision avoidance strategy. The control submodule is used to generate warning prompts and / or avoidance response instructions corresponding to the target collision avoidance strategy, and control the vehicle to execute the target collision avoidance strategy according to the warning prompts and / or avoidance response instructions.

[0072] Optionally, the third determining submodule includes: The first processing unit is configured to, if the collision risk level is the first level, include a visual warning prompt as the target anti-collision strategy corresponding to the first level. The second processing unit is configured to, if the collision risk level is a second level which is higher than the first level, then the target anti-collision strategy corresponding to the second level includes visual warning prompts and audible warning prompts. The third processing unit is used to, if the collision risk level is a third level which is higher than the second level, then the target collision avoidance strategy corresponding to the third level includes visual warning prompts, sound warning prompts, steering wheel vibration prompts, and hazard avoidance reminders. The fourth processing unit is used to determine the target collision avoidance strategy corresponding to the fourth level if the collision risk level is the fourth level, which is higher than the third level. This strategy includes visual warning prompts, audible warning prompts, steering wheel vibration prompts, and active avoidance.

[0073] Optionally, the device further includes: The lane change determination module is used to determine the active lane change to be performed by the vehicle when the collision avoidance strategy of the target includes active avoidance. The parameter determination module is used to monitor whether the vehicle is in an authorized state of active risk avoidance and whether the current environment of the vehicle meets the conditions for lane changing, and to determine the active lane changing parameters; the active lane changing parameters include the target lane, the lateral angular velocity of the vehicle, and the steering angle; The lane change module is used to control the vehicle to perform active lane changes according to the active lane change parameters.

[0074] The vehicle collision avoidance device provided in this application embodiment can realize each process of the vehicle collision avoidance method in the above embodiments of this application. To avoid repetition, it will not be described again here.

[0075] The vehicle collision avoidance device provided in this application acquires target vehicle information and lane information behind the vehicle based on heterogeneous sensors, including a visual camera, millimeter-wave radar, and lidar. It acquires vehicle information and, based on the vehicle information, target vehicle information, and lane information, identifies obstacle vehicles from the target vehicles behind the vehicle. Based on the relative operating states of the vehicle and the obstacle vehicles, it determines the collision risk level between the vehicle and the obstacle vehicles. According to a preset graded collision avoidance strategy, it controls the vehicle to execute a target collision avoidance strategy corresponding to the collision risk level. The graded collision avoidance strategy includes warning prompts and / or hazard avoidance responses. This application embodiment utilizes multiple heterogeneous sensors to achieve comprehensive and accurate perception of the environment behind the vehicle, overcoming the limitations of a single sensor in complex environments. By employing information from the vehicle itself, the target vehicle, and lane information, it can quickly and accurately identify obstructing vehicles behind without significantly increasing costs. This provides a reliable data foundation for autonomous risk avoidance assisted driving. By comprehensively considering the relative operating states of the vehicle and obstructing vehicles, such as relative speed, relative distance, approach time, and road conditions, a multi-dimensional risk assessment is conducted. This enables precise collision risk level classification and warning timing judgment, and provides graded collision avoidance warnings and interventions based on the collision risk level. This allows the vehicle to actively and effectively avoid dangerous vehicles behind, further improving vehicle driving safety.

[0076] Reference Figure 4 This application also provides an electronic device, such as... Figure 4 As shown, it includes a processor 301, a communication interface 302, a memory 303, and a communication bus 304, wherein the processor 301, the communication interface 302, and the memory 303 communicate with each other through the communication bus 304. Processor 301, memory 303 for storing processor-executable instructions; The processor 301 is configured to execute the instructions to implement the vehicle collision avoidance method described above.

[0077] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0078] The communication interface is used for communication between the aforementioned terminal and other devices.

[0079] The memory may include random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0080] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0081] In another embodiment provided in this application, a computer-readable storage medium is also provided, on which a computer program is stored, which, when executed by a processor, implements any of the vehicle collision avoidance methods described in the above embodiments.

[0082] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0083] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0084] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0085] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application are included within the scope of protection of this application.

Claims

1. A vehicle collision avoidance method, characterized in that, The method includes: Acquire target vehicle information and lane information behind the vehicle based on heterogeneous sensors; the heterogeneous sensors include a visual camera, millimeter-wave radar, and lidar. Acquire vehicle information, and based on the vehicle information, the target vehicle information, and the lane information, identify the obstacle vehicle from the target vehicles behind the vehicle. The collision risk level between the vehicle and the obstacle vehicle is determined based on their relative operating states. According to the preset graded collision avoidance strategy, the vehicle is controlled to execute the target collision avoidance strategy corresponding to the collision risk level, wherein the graded collision avoidance strategy includes early warning prompts and / or avoidance responses.

2. The method according to claim 1, characterized in that, The acquisition of target vehicle information and lane information behind the vehicle based on heterogeneous sensors includes: Acquire environmental data behind the vehicle collected by various heterogeneous sensors; Based on the accuracy information of each heterogeneous sensor in the current environment, determine the weighting coefficients of the same environmental data collected by each heterogeneous sensor. Based on the weighting coefficients, the same environmental data collected by each heterogeneous sensor is weighted and fused to obtain target vehicle information and lane information behind the vehicle.

3. The method according to claim 1, characterized in that, The step of acquiring vehicle information and identifying obstacle vehicles from target vehicles behind the vehicle based on the vehicle information, the target vehicle information, and the lane information includes: Obtain vehicle information, and generate the running trajectories of the vehicle and the target vehicle within the historical period based on the vehicle information and the target vehicle information, respectively. Based on the lane information, extract the geometric features of the lane where the vehicle is currently located; Based on the historical trajectories of the autonomous vehicle and the target vehicle, as well as the geometric features of the lane currently occupied by the autonomous vehicle, obstacle vehicles are identified from the target vehicles behind the autonomous vehicle.

4. The method according to claim 1, characterized in that, The method of determining the collision risk level between the vehicle and the obstacle vehicle based on their relative operating states includes: Based on the vehicle information and the target vehicle information corresponding to the obstacle vehicle, the relative speed, relative acceleration, and relative distance between the vehicle and the obstacle vehicle are calculated. The collision time between the vehicle and the obstacle vehicle is calculated based on the relative speed, relative acceleration, and relative distance. Based on the relative speed and the collision time, the relative operating state of the vehicle and the obstacle vehicle is determined, and the relative operating state is compared with the preset judgment conditions to determine the collision risk level between the vehicle and the obstacle vehicle.

5. The method according to claim 1, characterized in that, The step of controlling the vehicle to execute a target collision avoidance strategy corresponding to the collision risk level according to a preset graded collision avoidance strategy includes: Obtain each collision risk level and its corresponding preset graded collision avoidance strategy; Based on the preset graded collision avoidance strategy, determine the target collision avoidance strategy corresponding to the collision risk level; Generate warning prompts and / or hazard avoidance response commands corresponding to the target collision avoidance strategy, and control the vehicle to execute the target collision avoidance strategy according to the warning prompts and / or hazard avoidance response commands.

6. The method according to claim 5, characterized in that, The step of determining the target collision avoidance strategy corresponding to the collision risk level based on the preset graded collision avoidance strategy includes: If the collision risk level is the first level, then the target collision avoidance strategy corresponding to the first level includes visual warning prompts; If the collision risk level is a second level, which is higher than the first level, then the target collision avoidance strategy corresponding to the second level includes visual warning prompts and audible warning prompts. If the collision risk level is a third level, which is higher than the second level, then the target collision avoidance strategy corresponding to the third level includes visual warning prompts, audible warning prompts, steering wheel vibration prompts, and hazard avoidance reminders. If the collision risk level is level four, which is higher than level three, then the target collision avoidance strategy corresponding to level four includes visual warning prompts, audible warning prompts, steering wheel vibration prompts, and active avoidance.

7. The method according to claim 1, characterized in that, The method further includes: When the collision avoidance strategy of the target includes active avoidance, it is determined that the vehicle needs to perform an active lane change; The system detects that the vehicle is in an authorized state for active risk avoidance and that the current environment of the vehicle meets the conditions for lane changing, and determines the active lane changing parameters; the active lane changing parameters include the target lane, the lateral angular velocity of the vehicle, and the steering angle; The vehicle is controlled to perform an active lane change according to the aforementioned active lane change parameters.

8. A vehicle collision avoidance device, characterized in that, The device includes: The information acquisition module is used to acquire target vehicle information and lane information behind the vehicle based on heterogeneous sensors; the heterogeneous sensors include a visual camera, millimeter-wave radar and lidar. An obstacle recognition module is used to acquire vehicle information and, based on the vehicle information, the target vehicle information, and the lane information, identify obstacle vehicles from the target vehicles behind the vehicle. The risk assessment module is used to determine the collision risk level between the vehicle and the obstacle vehicle based on the relative operating status of the vehicle and the obstacle vehicle. The control module is used to control the vehicle to execute a target collision avoidance strategy corresponding to the collision risk level according to a preset graded collision avoidance strategy, wherein the graded collision avoidance strategy includes early warning prompts and / or hazard avoidance responses.

9. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to execute the instructions to implement the vehicle collision avoidance method as described in any one of claims 1 to 7.

10. A readable storage medium, characterized in that, A computer program is stored on the readable storage medium, which, when executed by a processor, implements the vehicle collision avoidance method as described in any one of claims 1 to 7.

Images