Vehicle emergency braking system and its decision control method and device

Abstract

This invention discloses a vehicle emergency braking system and its decision control method and apparatus. The decision control method includes: acquiring obstacle information through a binocular stereo camera and millimeter-wave radar; determining a first dynamic weight for the first data acquired by the binocular stereo camera and a second dynamic weight for the second data acquired by the millimeter-wave radar, based on the current obstacle type, environmental parameters, and distance level; weightedly fusing one or more obstacle-related parameters according to the first data, the second data, the first dynamic weight, and the second dynamic weight to obtain fused data; obtaining a decision processing result based on the fused data and the vehicle status data synchronously transmitted via the vehicle's CAN bus; and determining the output control command based on the decision processing result. Using this scheme can improve perception accuracy, enhance the accuracy of emergency braking, effectively reduce the likelihood of traffic accidents, and ensure vehicle driving safety.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving technology, and more specifically, to a vehicle emergency braking system and its decision control method and apparatus. Background Technology

[0002] With increased attention to traffic safety and the implementation of relevant standards, the Automatic Emergency Braking System (AEBS) has become a core piece of equipment for ensuring driving safety. In particular, passenger vehicles and dangerous goods vehicles are required to install intelligent safety equipment that meets the standards.

[0003] Existing AEBS (Autonomous Emergency Braking System) perception modules mostly employ single or combined sensor solutions such as monocular cameras, millimeter-wave radar, and lidar. However, these solutions have significant drawbacks: monocular cameras can only identify vehicles and pedestrians, failing to detect other types of obstacles; millimeter-wave radar has weak pedestrian detection capabilities and is easily interfered with by metallic objects; lidar is costly, has poor penetration, and performs poorly in detecting black objects; and ultrasonic radar has a short detection range, making it only suitable for reversing scenarios. These shortcomings result in insufficient perception accuracy of existing AEBS systems in complex road conditions, making it difficult to comprehensively cover various collision risks.

[0004] Therefore, providing a vehicle emergency braking solution to improve perception accuracy, thereby enhancing the accuracy of emergency braking, effectively reducing the likelihood of traffic accidents, and ensuring vehicle driving safety is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] The main objective of this invention is to disclose a vehicle emergency braking system and its decision control method and device, so as to at least solve the problems of insufficient perception accuracy of AEBS system in complex road conditions and difficulty in fully covering various collision risks in related technologies.

[0006] According to one aspect of the present invention, a decision control method based on a vehicle emergency braking system is provided.

[0007] The decision control method based on a vehicle emergency braking system provided by the present invention includes: acquiring obstacle information through a binocular stereo camera and a millimeter-wave radar; determining a first dynamic weight of the first data acquired by the binocular stereo camera and a second dynamic weight of the second data acquired by the millimeter-wave radar by combining obstacle type, environmental parameters and distance level; performing weighted fusion on one or more obstacle-related parameters according to the first data, the second data, the first dynamic weight and the second dynamic weight to obtain fused data; obtaining a decision processing result according to the fused data and the vehicle status data synchronously transmitted by the vehicle CAN bus, and determining the output control command according to the decision processing result.

[0008] The determination of the first dynamic weight of the first data acquired by the stereo camera and the second dynamic weight of the second data acquired by the millimeter-wave radar, based on the current obstacle type, environmental parameters, and distance level, may further include: determining the type of the current obstacle using the image semantic recognition algorithm of the stereo camera, where each obstacle type corresponds to a weight value ω1; determining the weight value ω2 corresponding to the current environmental parameters by combining data from the vehicle-mounted light sensor and rain sensor; determining the current distance level to which the current detection distance belongs, where each distance level corresponds to a weight value ω3; determining the first dynamic weight of the first data acquired by the stereo camera based on the weight value ω1 corresponding to the current obstacle type, the weight value ω2 corresponding to the current environmental parameters, and the weight value ω3 corresponding to the current distance level, and determining the second dynamic weight of the second data acquired by the millimeter-wave radar based on the first dynamic weight.

[0009] Based on the weight value ω1 corresponding to the current obstacle type, the weight value ω2 corresponding to the current environmental parameters, and the weight value ω3 corresponding to the current distance level, the first dynamic weight of the first data acquired by the binocular stereo camera is determined, and the second dynamic weight of the second data acquired by the millimeter-wave radar is determined based on the first dynamic weight, including: The first dynamic weight of the aforementioned first data is determined in the following manner. : = ω1×ω2×ω3 + k, where k is the calibration coefficient; The second dynamic weight of the aforementioned second data is determined in the following manner. : = 1 - .

[0010] Based on the aforementioned first data, second data, first dynamic weight, and second dynamic weight, one or more obstacle-related parameters are weighted and fused to obtain the fused data, including: The distance parameters of obstacles are weighted and fused using the following method to obtain the fused data. : = × + × ; The velocity parameters of the obstacles are weighted and fused using the following method to obtain the fused data. : = × + × ; in, For distance measurements based on binocular stereo cameras, The velocity measurements are based on a stereo camera. For distance measurements based on millimeter-wave radar, The velocity measurement value is from millimeter-wave radar. The first dynamic weight of the aforementioned first data. This is the second dynamic weight of the second data mentioned above.

[0011] Based on the fused data and the vehicle status data transmitted via the vehicle's CAN bus, a decision processing result is obtained, and the output control commands are determined based on the decision processing result, including: predicting the longitudinal distance of obstacles within a predetermined time based on the fused data. Speed ​​of obstacles Based on the longitudinal distance of the aforementioned obstacles Speed ​​of obstacles Determine the dynamic collision time of moving obstacles Dynamic collision time with stationary obstacles Based on road conditions and the aforementioned vehicle status data, the dynamic alarm threshold is calculated. With braking threshold ;when < ,or, < When, send a braking command; when ≤ < ,or, ≤ < When the vehicle is in a braking state or in a non-forward gear, an alarm command is sent; when the vehicle is in a braking state or in a non-forward gear, no command is sent.

[0012] Based on the above longitudinal distance of the obstacles Speed ​​of obstacles Determine the dynamic collision time of moving obstacles Dynamic collision time with stationary obstacles include: The dynamic collision time of a moving obstacle is determined using the following methods. : = / ( - ) ; Determine the dynamic collision time of a stationary obstacle using the following method. : = / .

[0013] Among them, the above This refers to the vehicle's speed.

[0014] Based on road conditions and the aforementioned vehicle status data, the dynamic alarm threshold is calculated. With braking threshold include: The dynamic alarm threshold is calculated using the following method. : ; The braking threshold is calculated in the following way. : ; Where μ is the road adhesion coefficient, M is the load, M0 is the standard vehicle load, and a, b, c, d, e, and f are calibration parameters that satisfy the constraints. < .

[0015] After determining the output control command based on the above decision-making results, it also includes at least one of the following: determining the adhesion coefficient level corresponding to the current road surface adhesion coefficient, adapting the corresponding braking strategy according to the adhesion coefficient level, and determining the corresponding braking parameters; based on the braking threshold... Dynamic collision time of moving obstacles and dynamic collision time of stationary obstacles Determine the collision risk level and adapt the corresponding early warning method according to the collision risk level.

[0016] According to another aspect of the present invention, a decision control device based on a vehicle emergency braking system is provided.

[0017] The decision control device based on a vehicle emergency braking system according to the present invention includes: a data acquisition module for acquiring obstacle information via a binocular stereo camera and a millimeter-wave radar; a determination module for determining a first dynamic weight of the first data acquired by the binocular stereo camera and a second dynamic weight of the second data acquired by the millimeter-wave radar, based on the current obstacle type, environmental parameters, and distance level; an acquisition module for weighted fusion of one or more obstacle-related parameters according to the first data, the second data, the first dynamic weight, and the second dynamic weight, to acquire fused data; and a decision control module for obtaining a decision processing result based on the fused data and vehicle status data transmitted via the vehicle CAN bus, and determining an output control command based on the decision processing result.

[0018] According to another aspect of the present invention, a vehicle emergency braking system is provided.

[0019] The vehicle emergency braking system according to the present invention includes: a perception module, a decision controller, an execution module, and a vehicle CAN bus. The perception module includes: a binocular stereo camera and a millimeter-wave radar, used to collect obstacle information; the vehicle CAN bus is used to synchronously transmit vehicle status data; the decision controller is used to determine a first dynamic weight of the first data collected by the binocular stereo camera and a second dynamic weight of the second data collected by the millimeter-wave radar, based on the current obstacle type, environmental parameters, and distance level; to perform weighted fusion of one or more obstacle-related parameters according to the first data, the second data, the first dynamic weight, and the second dynamic weight, to obtain fused data; to obtain a decision processing result based on the fused data and the vehicle status data synchronously transmitted by the vehicle CAN bus; and to determine the output control command based on the decision processing result; the execution module is used to execute audible and visual warnings or graded braking operations after receiving the control command from the decision controller.

[0020] The vehicle emergency braking system and its decision control method and device provided by this invention acquire obstacle information through a binocular stereo camera and millimeter-wave radar; combining obstacle type, environmental parameters, and distance level, a first dynamic weight of the first data acquired by the binocular stereo camera and a second dynamic weight of the second data acquired by the millimeter-wave radar are determined; based on the first data, the second data, the first dynamic weight, and the second dynamic weight, one or more obstacle-related parameters are weighted and fused to obtain fused data. Based on the fused data and the vehicle status data synchronously transmitted via the vehicle's CAN bus, a decision processing result is obtained, and the output control command is determined based on the decision processing result. Since the binocular stereo camera simulates the distance measurement principle of the human eye, it can acquire three-dimensional point cloud data of all obstacles in the field of view through the triangular relationship of parameters such as parallax and focal length, possessing the advantages of high distance measurement accuracy and the ability to detect all types of obstacles. Integrating binocular stereo cameras with millimeter-wave radar into an automatic emergency braking system (AEBS) can overcome the limitations of a single sensor. By considering dynamic factors such as the current environment, obstacle type, and distance, and optimizing the sensor fusion algorithm, it can achieve accurate detection of various obstacles, thereby improving perception accuracy, enhancing the accuracy of emergency braking, effectively reducing the likelihood of traffic accidents, and ensuring vehicle driving safety. Attached Figure Description

[0021] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0022] Figure 1 This is a flowchart of a decision control method based on a vehicle emergency braking system according to an embodiment of the present invention; Figure 2 This is a structural block diagram of a decision control device based on a vehicle emergency braking system according to an embodiment of the present invention; Figure 3 This is a structural block diagram of a vehicle emergency braking system according to an embodiment of the present invention. Detailed Implementation

[0023] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] The specific implementation of the present invention will now be described in detail with reference to the accompanying drawings.

[0025] According to an embodiment of the present invention, a decision control method based on a vehicle emergency braking system is provided.

[0026] Figure 1 This is a flowchart of a decision-making and control method based on a vehicle emergency braking system according to an embodiment of the present invention. Figure 1 As shown, the decision control method based on the vehicle emergency braking system includes: Step S101: Obstacle information is acquired using a binocular stereo camera and millimeter-wave radar; Step S102: Combine the current obstacle type, environmental parameters and distance level to determine the first dynamic weight of the first data collected by the stereo camera and the second dynamic weight of the second data collected by the millimeter-wave radar. Step S103: Based on the first data, the second data, the first dynamic weight, and the second dynamic weight, perform weighted fusion on one or more parameters related to the obstacle to obtain fused data; Step S104: Based on the above-mentioned fused data and the vehicle status data synchronously transmitted by the vehicle CAN bus, obtain the decision processing result, and determine the output control command based on the above-mentioned decision processing result.

[0027] use Figure 1 The method shown integrates a binocular stereo camera with millimeter-wave radar in an automatic emergency braking system (AEBS). This approach compensates for the limitations of a single sensor and takes into account dynamic factors such as the current environment, obstacle type, and distance. By optimizing the sensor fusion algorithm, it achieves accurate detection of various obstacles, thereby improving perception accuracy, enhancing the accuracy of emergency braking, effectively reducing the likelihood of traffic accidents, and ensuring vehicle driving safety.

[0028] The vehicle emergency braking system provided in this application includes: a perception module, a decision controller, an execution module, and a vehicle CAN bus. The perception module includes, but is not limited to, a binocular stereo camera and millimeter-wave radar. Its core innovation lies in employing a dynamic weight fusion algorithm based on "obstacle type + environmental parameters," replacing the fixed-distance weight allocation method, to improve perception accuracy and anti-interference capabilities. The perception module collects obstacle information in real time, and the vehicle CAN bus synchronously transmits vehicle status data; both are sent to the decision controller. The decision controller outputs corresponding alarm or braking commands based on the fusion algorithm and risk assessment logic. The execution module includes a braking execution system and a human-machine interface screen. After receiving commands, the execution module achieves risk avoidance through audible and visual warnings or graded braking.

[0029] In step S102 above, determining the first dynamic weight of the first data acquired by the binocular stereo camera, based on the current obstacle type, environmental parameters, and distance level, and determining the second dynamic weight of the second data acquired by the millimeter-wave radar, may further include the following processing: Step 1.1: Determine the type of the current obstacle using the image semantic recognition algorithm of the stereo camera, where each obstacle type corresponds to a weight value ω1; Step 1.2: Combine the data from the vehicle's light sensor and rain sensor to determine the weight value ω2 corresponding to the current environmental parameters; Step 1.3: Determine the current distance level to which the current detection distance belongs, where each distance level corresponds to a weight value ω3; Step 1.4: Based on the weight value ω1 corresponding to the current obstacle type, the weight value ω2 corresponding to the current environmental parameter, and the weight value ω3 corresponding to the current distance level, determine the first dynamic weight of the first data collected by the binocular stereo camera, and determine the second dynamic weight of the second data collected by the millimeter-wave radar based on the first dynamic weight.

[0030] In step 1.4 above, the first dynamic weight of the first data acquired by the stereo camera is determined based on the weight value ω1 corresponding to the current obstacle type, the weight value ω2 corresponding to the current environmental parameters, and the weight value ω3 corresponding to the current distance level. The second dynamic weight of the second data acquired by the millimeter-wave radar is then determined based on the first dynamic weight, including: The first dynamic weight of the aforementioned first data is determined in the following manner. : = ω1×ω2×ω3 + k, where k is the calibration coefficient; The second dynamic weight of the aforementioned second data is determined in the following manner. : = 1 - .

[0031] In step S103, based on the first data, the second data, the first dynamic weight, and the second dynamic weight, one or more parameters related to the obstacle are weighted and fused to obtain the fused data. This process may further include the following steps: The distance parameters of obstacles are weighted and fused using the following method to obtain the fused data. : = × + × ; The velocity parameters of the obstacles are weighted and fused using the following method to obtain the fused data. : = × + × ; in, For distance measurements based on binocular stereo cameras, The velocity measurements are based on a stereo camera. For distance measurements based on millimeter-wave radar, The velocity measurement value is from millimeter-wave radar. The first dynamic weight of the aforementioned first data. This is the second dynamic weight of the second data mentioned above.

[0032] In the preferred implementation process, the perception module includes a binocular stereo camera and a millimeter-wave radar. The core innovation lies in the dynamic weight fusion algorithm of "obstacle type + environmental parameters", which replaces the traditional fixed distance weight allocation method and improves perception accuracy and anti-interference capability.

[0033] (1) Definition of core parameters Obstacle type factor (ω1): Obstacles can be classified into five categories through image semantic recognition by a stereo camera: pedestrians, non-motorized vehicles, large vehicles, small vehicles, and irregular static obstacles (such as scattered goods). ω1 is assigned values ​​of 0.8, 0.7, 0.5, 0.6, and 0.9 respectively (the larger the value, the higher the weight of the stereo camera data). Environmental parameter factor (ω2): Combining data from the vehicle's light sensor and rain sensor, it is quantified into a value between 0 and 1 (for example, when the light is too strong or the humidity is high in rainy weather, ω2 can be reduced to increase the radar weight). Distance level factor (ω3): Divided into short distance (e.g., 0~5m range, ω3 can be 0.3), medium distance (e.g., 5~100m range, ω3 can be 0.8), and long distance (e.g., >100m range, ω3 can be 0.4) according to the detection distance, which characterizes the adaptability of the binocular camera at different distances.

[0034] (2) Dynamic weight calculation Define dynamic weights for binocular camera data Dynamic weighting of millimeter-wave radar data (Constraints:) + = 1), the calculation formula simplifies to: = ω1 × ω2 × ω3 + k = 1 -

[0035] Where k is the calibration coefficient (0.05-0.1), to avoid extreme weighting caused by a single factor.

[0036] (3) Data fusion output For the core parameters of the obstacle (distance S, velocity) We perform weighted fusion to output accurate perception results: = × + ×

[0037] = × + ×

[0038] in, , These are measurements taken from a stereo camera. , These are measurements taken by millimeter-wave radar.

[0039] In step S104, based on the fused data and the vehicle status data transmitted via the vehicle CAN bus, a decision processing result is obtained, and the output control command determined based on the decision processing result may further include the following processing: Step 2.1: Based on the fused data above, predict the longitudinal distance of obstacles within a predetermined time. Speed ​​of obstacles ; Step 2.2: Based on the longitudinal distance of the obstacles mentioned above Speed ​​of obstacles Determine the dynamic collision time of moving obstacles Dynamic collision time with stationary obstacles ; Step 2.3: Based on the road conditions and the above vehicle status data, calculate the dynamic alarm threshold. With braking threshold ; Step 2.4: When < ,or, < At that time, a braking command is sent; when ≤ < ,or, ≤ < When the vehicle is in a braking state or in a non-forward gear, an alarm command is sent; when the vehicle is in a braking state or in a non-forward gear, no command is sent (to avoid interfering with manual operation).

[0040] In step 2.1, for example, a simplified LSTM trajectory prediction model can be used, taking into input obstacle fusion data (lateral distance, longitudinal distance, velocity) from multiple past frames (e.g., 5 frames, 100ms per frame) to predict the longitudinal distance of obstacles within a predetermined time period in the future (e.g., within 300ms). ) and speed ( This avoids the decision lag caused by the traditional "uniform velocity assumption." The calculation process of the LSTM trajectory prediction model can be found in the description of related technologies, and will not be repeated here.

[0041] In step 2.2, based on the aforementioned longitudinal distance of the obstacle... Speed ​​of obstacles Determine the dynamic collision time of moving obstacles Dynamic collision time with stationary obstacles The following processing may be further included: The dynamic collision time of a moving obstacle is determined using the following methods. : = / ( - ) ; Determine the dynamic collision time of a stationary obstacle using the following method. : = / .

[0042] Among them, the above This refers to the vehicle's speed.

[0043] In the preferred implementation process, based on the above predicted data, the collision time calculation formula is revised to obtain a risk assessment value that is more realistic: Dynamic collision time of moving obstacles : = / ( - ); ≥ hour, = ∞) Dynamic collision time of stationary obstacles : Right now =0, then = /

[0044] In step 2.3, the dynamic alarm threshold is calculated by combining the road conditions and the aforementioned vehicle status data. With braking threshold This may include the following processing: The dynamic alarm threshold is calculated using the following method. : ; The braking threshold is calculated in the following way. : ; Where μ is the road adhesion coefficient, M is the load, M0 is the standard vehicle load, and a, b, c, d, e, and f are calibration parameters that satisfy the constraints. < .

[0045] In the preferred implementation process, this application does not adopt a fixed threshold scheme, but instead constructs a dynamic alarm threshold by combining the vehicle status and road conditions. With braking threshold The calculation formula simplifies to: ; ; Where μ is the road surface adhesion coefficient (simplified using onboard sensors), and M is the load. M0 represents the vehicle's standard load, and a, b, c, d, e, and f are calibration parameters (example values: a=0.8, b=0.3, c=0.5, d=0.5, e=0.2, f=0.2), satisfying the constraints. < .

[0046] After step S104, and after determining the output control command based on the above decision processing result, it may further include at least one of the following: Determine the adhesion coefficient level corresponding to the current road surface adhesion coefficient, adapt the corresponding braking strategy according to the adhesion coefficient level, and determine the corresponding braking parameters. According to the braking threshold Dynamic collision time of moving obstacles and dynamic collision time of stationary obstacles Determine the collision risk level and adapt the corresponding early warning method according to the collision risk level.

[0047] In the preferred implementation process, the execution module may include, but is not limited to, the braking execution system and the human-machine interaction screen. The core innovation lies in the graded braking and multimodal warning based on the "road surface adhesion coefficient adaptation", which improves the safety and effectiveness of the response.

[0048] (1) Simplified estimation of road surface adhesion coefficient (μ) By using data from wheel speed sensors and longitudinal acceleration sensors, the estimation of μ value can be simplified. For example, μ can be divided into three levels: high adhesion (e.g., μ≥0.7, dry road), medium adhesion (e.g., 0.3≤μ<0.7, wet road), and low adhesion (e.g., μ<0.3, icy road).

[0049] (2) Graded braking strategy, as shown in Table 1:

[0050] (3) Multimodal early warning mechanism According to the braking threshold Dynamic collision time of moving obstacles and dynamic collision time of stationary obstacles Determine the collision risk level, for example, The risk level is determined based on the magnitude of ΔT, and then the warning method is adapted according to the collision risk level: For example, for low-risk levels (when ΔT≥2s), the warning method is: only the human-computer interaction screen icon prompt; Medium risk level (when 1s≤ΔT<2s), warning method: icon + voice reminder (adapted to road conditions, such as "slippery road surface, please slow down smoothly"); High risk level (when ΔT<1s), warning method: icon + rapid beep + steering wheel vibration, with simultaneous pre-load of braking pressure.

[0051] According to an embodiment of the present invention, a decision control device based on a vehicle emergency braking system is provided.

[0052] Figure 2 This is a structural block diagram of a decision control device based on a vehicle emergency braking system according to an embodiment of the present invention. Figure 2As shown, the decision control device based on the vehicle emergency braking system includes: a data acquisition module 20, used to acquire obstacle information through a binocular stereo camera and a millimeter-wave radar; a determination module 22, used to determine the first dynamic weight of the first data acquired by the binocular stereo camera and the second dynamic weight of the second data acquired by the millimeter-wave radar by combining the current obstacle type, environmental parameters and distance level; an acquisition module 24, used to perform weighted fusion of one or more obstacle-related parameters according to the first data, the second data, the first dynamic weight and the second dynamic weight to acquire fused data; and a decision control module 26, used to obtain a decision processing result based on the fused data and the vehicle status data transmitted by the vehicle CAN bus, and determine the output control command based on the decision processing result.

[0053] use Figure 2 The device shown integrates a binocular stereo camera with millimeter-wave radar in an automatic emergency braking system (AEBS). This overcomes the limitations of a single sensor and takes into account dynamic factors such as the current environment, obstacle type, and distance. By optimizing the sensor fusion algorithm, it achieves accurate detection of various obstacles, thereby improving perception accuracy, enhancing the accuracy of emergency braking, effectively reducing the likelihood of traffic accidents, and ensuring vehicle driving safety.

[0054] It should be noted that the aforementioned decision-making and control device based on the vehicle emergency braking system can be referred to in the corresponding document. Figure 1 The relevant descriptions and effects in the illustrated embodiments are for understanding purposes only and will not be repeated here.

[0055] According to an embodiment of the present invention, a vehicle emergency braking system is provided.

[0056] Figure 3 This is a structural block diagram of a vehicle emergency braking system according to an embodiment of the present invention. Figure 3As shown, the vehicle emergency braking system includes: a perception module 30, a decision controller 32, an execution module 34, and a vehicle CAN bus 36. The perception module 30 includes: a binocular stereo camera 300 and a millimeter-wave radar 302, which are used to collect obstacle information. The vehicle CAN bus 36 is used to synchronously transmit vehicle status data. The decision controller 32 is connected to the perception module 30 and the vehicle CAN bus 36, and is used to determine the first dynamic of the first data collected by the binocular stereo camera by combining the current obstacle type, environmental parameters, and distance level. The system calculates the weights and determines the second dynamic weights of the second data acquired by the millimeter-wave radar. Based on the first data, the second data, the first dynamic weights, and the second dynamic weights, it performs weighted fusion on one or more parameters related to the obstacle to obtain fused data. Based on the fused data and the vehicle status data synchronously transmitted via the vehicle CAN bus, it obtains the decision processing result and determines the output control command based on the decision processing result. The execution module 34, connected to the decision controller 32, is used to execute audible and visual warnings or graded braking operations after receiving the control command from the decision controller.

[0057] In the preferred implementation process, the automatic emergency braking system for vehicles provided in this application includes, but is not limited to: a perception module, a decision controller, an execution module, and a vehicle CAN bus. The above modules work together to achieve obstacle detection, risk assessment, and braking / alarm response.

[0058] The aforementioned perception module system further includes: a binocular stereo camera and a millimeter-wave radar; the aforementioned execution module further includes: a braking execution system and a human-machine interface screen. The aforementioned vehicle automatic emergency braking system may also include: transmission lines connecting the various components.

[0059] Workflow: The perception module (binocular stereo camera, millimeter-wave radar) collects obstacle information in real time, and the vehicle's CAN bus synchronously transmits the vehicle's status data. Both are sent to the decision controller. The decision controller outputs corresponding alarm or braking commands through the fusion algorithm and risk judgment logic. After receiving the command, the execution module executes the sound and light warning through the human-machine interaction screen, or executes the system's graded braking through the brakes to achieve risk avoidance.

[0060] It should be noted that the above-mentioned vehicle emergency braking systems can be found in the corresponding references. Figure 1 The relevant descriptions and effects in the illustrated embodiments are for understanding purposes only and will not be repeated here.

[0061] In summary, using the embodiments provided by this invention, a high-precision, intelligent, and highly adaptable AEBS (Automatic Emergency Braking System) is constructed through dynamic weight fusion of the perception module of the vehicle emergency braking system, trajectory prediction and dynamic thresholding of the decision controller, and hierarchical collaborative response of the execution module. The perception module includes a binocular stereo camera and millimeter-wave radar. Its core innovation lies in the dynamic weight fusion algorithm of "obstacle type + environmental parameters," which improves perception accuracy and anti-interference capability. The decision controller is the core of the vehicle's automatic emergency braking system, receiving fused data from the perception module and vehicle status data (vehicle speed) from the vehicle's CAN bus. The system assesses collision risk based on "trajectory prediction + dynamic threshold" logic, considering factors such as steering status, braking status, and gear position, and outputs precise commands. The execution module includes a braking execution system and a human-machine interface screen. Its core innovation lies in graded braking and multimodal warning based on "road surface adhesion coefficient adaptation," enhancing response safety and effectiveness. By optimizing sensor fusion algorithms, innovating decision-making logic, and upgrading the execution mechanism, the system achieves accurate detection, intelligent judgment, and graded response to various obstacles, thereby improving driving safety and system adaptability in complex scenarios. Compared to existing technologies, this application addresses core pain points such as inaccurate obstacle detection, delayed decision-making, and rigid execution in complex scenarios. It can effectively reduce the incidence and severity of forward collisions, and is particularly suitable for vehicles with extremely high active safety requirements, such as passenger vehicles and hazardous material transport vehicles, demonstrating significant technological advancement and practical application value.

[0062] The above specific embodiments further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made on the basis of the technical solution of the present invention should be included within the scope of protection of the present invention.

Claims

1. A decision-making and control method based on a vehicle emergency braking system, characterized in that, include: Obstacle information is collected using a binocular stereo camera and millimeter-wave radar; Based on the current obstacle type, environmental parameters, and distance level, the first dynamic weight of the first data acquired by the binocular stereo camera is determined, and the second dynamic weight of the second data acquired by the millimeter-wave radar is determined. Based on the first data, the second data, the first dynamic weight, and the second dynamic weight, one or more parameters related to the obstacle are weighted and fused to obtain the fused data; Based on the fused data and the vehicle status data synchronously transmitted via the vehicle's CAN bus, a decision processing result is obtained, and the output control command is determined based on the decision processing result.

2. The method according to claim 1, characterized in that, Based on the current obstacle type, environmental parameters, and distance level, the first dynamic weight of the first data acquired by the binocular stereo camera is determined, and the second dynamic weight of the second data acquired by the millimeter-wave radar is determined, including: The image semantic recognition algorithm of a stereo camera is used to determine the type of the current obstacle, where each obstacle type corresponds to a weight value ω1; By combining data from the vehicle's light sensor and rain sensor, the weight value ω2 corresponding to the current environmental parameters is determined; Determine the current distance level to which the current detection distance belongs, where each distance level corresponds to a weight value ω3; Based on the weight value ω1 corresponding to the current obstacle type, the weight value ω2 corresponding to the current environmental parameter, and the weight value ω3 corresponding to the current distance level, the first dynamic weight of the first data collected by the binocular stereo camera is determined, and the second dynamic weight of the second data collected by the millimeter-wave radar is determined based on the first dynamic weight.

3. The method according to claim 2, characterized in that, Based on the weight value ω1 corresponding to the current obstacle type, the weight value ω2 corresponding to the current environmental parameters, and the weight value ω3 corresponding to the current distance level, the first dynamic weight of the first data acquired by the binocular stereo camera is determined, and the second dynamic weight of the second data acquired by the millimeter-wave radar is determined based on the first dynamic weight, including: The first dynamic weight of the first data is determined in the following manner. : = ω1×ω2×ω3 + k, where k is the calibration coefficient; The second dynamic weight of the second data is determined in the following manner. : = 1 - 。 4. The method according to claim 1, characterized in that, Based on the first data, the second data, the first dynamic weight, and the second dynamic weight, one or more obstacle-related parameters are weighted and fused to obtain the fused data, including: The distance parameters of obstacles are weighted and fused using the following method to obtain the fused data. : = × + × ； The velocity parameters of the obstacles are weighted and fused using the following method to obtain the fused data. : = × + × ； in, For distance measurements based on binocular stereo cameras, The velocity measurements are based on a stereo camera. For distance measurements based on millimeter-wave radar, The velocity measurement value is from millimeter-wave radar. The first dynamic weight of the first data. This is the second dynamic weight of the second data.

5. The method according to claim 1, characterized in that, Based on the fused data and the vehicle status data transmitted via the vehicle's CAN bus, a decision processing result is obtained, and the output control commands are determined based on the decision processing result, including: Based on the fused data, predict the longitudinal distance of the obstacle within a predetermined time period. Speed ​​of obstacles ; Based on the longitudinal distance of the obstacle Speed ​​of obstacles Determine the dynamic collision time of moving obstacles Dynamic collision time with stationary obstacles ; By combining road conditions and the vehicle's status data, a dynamic alarm threshold is calculated. With braking threshold ; when < ,or, < At that time, a braking command is sent; when ≤ < ,or, ≤ < When necessary, send an alarm command; No commands are sent when the vehicle is already braking or in a non-forward gear.

6. The method according to claim 5, characterized in that, Based on the longitudinal distance of the obstacle Speed ​​of obstacles Determine the dynamic collision time of moving obstacles Dynamic collision time with stationary obstacles include: The dynamic collision time of a moving obstacle is determined using the following methods. : = / ( - ) ； Determine the dynamic collision time of a stationary obstacle using the following method. : = / ； Among them, the This refers to the vehicle's speed.

7. The method according to claim 5, characterized in that, By combining road conditions and the vehicle's status data, a dynamic alarm threshold is calculated. With braking threshold include: The dynamic alarm threshold is calculated using the following method. : ; The braking threshold is calculated in the following way. : ; Where μ is the road adhesion coefficient, M is the load, M0 is the standard vehicle load, and a, b, c, d, e, and f are calibration parameters that satisfy the constraints. < .

8. The method according to any one of claims 1 to 7, characterized in that, After determining the output control command based on the decision processing result, it also includes at least one of the following: Determine the adhesion coefficient level corresponding to the current road surface adhesion coefficient, adapt the corresponding braking strategy according to the adhesion coefficient level, and determine the corresponding braking parameters; According to the braking threshold Dynamic collision time of moving obstacles and dynamic collision time of stationary obstacles Determine the collision risk level and adapt the corresponding warning method according to the collision risk level.

9. A decision control device based on a vehicle emergency braking system, characterized in that, include: The data acquisition module is used to acquire obstacle information using a binocular stereo camera and millimeter-wave radar. The determination module is used to determine the first dynamic weight of the first data acquired by the binocular stereo camera and the second dynamic weight of the second data acquired by the millimeter-wave radar, by combining the current obstacle type, environmental parameters and distance level. The acquisition module is used to perform weighted fusion of one or more parameters related to the obstacle based on the first data, the second data, the first dynamic weight, and the second dynamic weight, and to acquire the fused data; The decision control module is used to obtain decision processing results based on the fused data and the vehicle status data transmitted via the vehicle CAN bus, and to determine the output control commands based on the decision processing results.

10. A vehicle emergency braking system, characterized in that, include: The system includes a perception module, a decision controller, an execution module, and a vehicle CAN bus. The perception module includes a binocular stereo camera and a millimeter-wave radar, which are used to collect obstacle information. The vehicle CAN bus is used for synchronous transmission of vehicle status data; The decision controller is used to determine the first dynamic weight of the first data collected by the binocular stereo camera and the second dynamic weight of the second data collected by the millimeter-wave radar by combining the current obstacle type, environmental parameters and distance level. Based on the first data, the second data, the first dynamic weight and the second dynamic weight, it performs weighted fusion on one or more parameters related to the obstacle to obtain fused data. Based on the fused data and the vehicle status data synchronously transmitted by the vehicle CAN bus, it obtains the decision processing result and determines the output control command based on the decision processing result. The execution module is used to perform audible and visual warnings or graded braking operations after receiving control commands from the decision controller.

Images