A control method of vehicle automatic emergency braking, electronic equipment and vehicle

Abstract

The application discloses a kind of vehicle automatic emergency braking control method, electronic equipment and vehicle, the control method includes: the video data, laser radar point cloud and millimeter wave radar point cloud collected at the same time are acquired;Video detection target and its first state information are acquired based on the video data, laser radar detection target and its second state information are acquired based on the laser radar point cloud, millimeter wave radar detection target and its third state information are acquired based on the millimeter wave radar point cloud;The laser radar detection target and its second state information, the millimeter wave radar detection target and its third state information and the video detection target and its first state information are matched with each other and associated with vehicle predicted trajectory, to obtain common detection target and its state information;Vehicle emergency braking control strategy is executed based on the common detection target and its state information.Improves the detection precision of the state information of common detection target, avoids appearing false positive and false negative detection result, improves the performance of automatic emergency braking.

Description

Technical Field

[0001] This invention relates to the field of vehicle safety technology, and more specifically to a control method, electronic equipment, and vehicle for automatic emergency braking of a vehicle. Background Technology

[0002] With the development of intelligent vehicles, the adoption rate of intelligent driving systems in newly launched cars is also increasing. Vehicles equipped with Automatic Emergency Braking (AEB) can effectively avoid rear-end collisions, reduce the rate of fatal collisions, and decrease injuries in driving scenarios.

[0003] Existing AEB (Autonomous Emergency Braking) solutions are generally based on a 1R1V system, consisting of one front-mounted millimeter-wave radar and one camera. While AEB systems need to make accurate braking maneuvers at the right time, mis-identification of dangerous targets can lead to false braking and even greater potential risks, posing a significant challenge to AEB system providers. In a 1R1V AEB system, the radar detects the target's distance, speed, and azimuth, assigning each characteristic an ID. The camera, based on visual SLAM principles, detects the target's distance, speed, azimuth, type, height, width, and current depth, also assigning each characteristic an ID. Within this 1R1V system, one of the sensors collects raw data or tracking data from the other, performs signal processing, data processing, and function implementation, and outputs the results to the V-CAN bus for longitudinal vehicle control. During this process, the 1R1V system will only recognize the target as a valid target if the ID from the Radar is successfully matched with the ID from the camera. A single sensor's perception result will not be sufficient for the system to perform longitudinal control of the vehicle.

[0004] The drawback of the 1R1V system is that while the radar is highly accurate in detecting target distance and velocity, its accuracy in detecting target type, outline, and azimuth is very low, and it detects many false targets. The camera is highly accurate in detecting target type, outline, and azimuth, but its accuracy in detecting target distance and velocity is very low. When the 1R1V system performs target matching, it needs to match the detection data from the radar and camera. However, because the overlapping areas of their respective strengths and weaknesses in target detection are misaligned, target matching often fails. Ultimately, the 1R1V output results in false positives and false negatives, posing a significant threat to the driver's life. Summary of the Invention

[0005] The summary section introduces a series of simplified concepts, which will be further explained in detail in the detailed description section. The summary section of this invention is not intended to limit the key features and essential technical features of the claimed technical solution, nor is it intended to determine the scope of protection of the claimed technical solution.

[0006] This invention provides a control method for automatic emergency braking of a vehicle, characterized by comprising:

[0007] Acquire video data, lidar point cloud, and millimeter-wave radar point cloud collected at the same time;

[0008] Based on the video data, obtain the video detection target and its first state information; based on the lidar point cloud, obtain the lidar detection target and its second state information; based on the millimeter-wave radar point cloud, obtain the millimeter-wave radar detection target and its third state information.

[0009] The target detected by the lidar and its second state information, the target detected by the millimeter-wave radar and its third state information, and the target detected by the video and its first state information are matched with each other and associated with the vehicle's predicted trajectory to obtain a common target detected and its state information.

[0010] The vehicle emergency braking control strategy is executed based on the common detection target and its status information.

[0011] Furthermore, obtaining video detection targets and their first state information based on the video data, obtaining lidar detection targets and their second state information based on the lidar point cloud, and obtaining millimeter-wave radar detection targets and their third state information based on the millimeter-wave radar point cloud include:

[0012] Based on the video data, obtain one or more video detection targets and their first state information;

[0013] Based on the lidar point cloud, acquire one or more lidar-detected targets and their second state information;

[0014] Based on the millimeter-wave radar point cloud, one or more millimeter-wave radar-detected targets and their third state information are obtained.

[0015] Further, matching the target detected by the lidar and its second state information, the target detected by the millimeter-wave radar and its third state information, and the target detected by the video and its first state information with each other and associating them with the vehicle's predicted trajectory to obtain common detection target and its state information includes the following steps:

[0016] The one or more lidar-detected targets and their second state information are matched with the one or more millimeter-wave radar-detected targets and their third state information to obtain a pair of lidar-detected targets and millimeter-wave radar-detected targets, wherein the pair of lidar-detected targets and millimeter-wave radar-detected targets are the same object.

[0017] The one or more lidar-detected targets and their second state information are matched with the one or more video-detected targets and their first state information to obtain a pair of lidar-detected targets and video-detected targets, wherein the pair of lidar-detected targets and video-detected targets are the same object;

[0018] Based on the pair of lidar-detected targets and millimeter-wave radar-detected targets, and the pair of lidar-detected targets and video-detected targets, a set of lidar-detected targets, millimeter-wave radar-detected targets, and video-detected targets are obtained;

[0019] The set of LiDAR-detected targets, millimeter-wave radar-detected targets, and video-detected targets are associated with the vehicle's predicted trajectory to obtain common detection targets and their status information.

[0020] Furthermore, associating the set of lidar-detected targets, millimeter-wave radar-detected targets, and video-detected targets with the vehicle's predicted trajectory to obtain common detected targets and their state information includes:

[0021] Obtain a first distance between the target detected by the lidar and the predicted trajectory of the vehicle; obtain a second distance between the target detected by the millimeter-wave radar and the predicted trajectory of the vehicle; obtain a third distance between the target detected by the video and the predicted trajectory of the vehicle.

[0022] When the difference between any two of the first distance, the second distance, and the third distance is less than a threshold, the group of LiDAR-detected targets, millimeter-wave radar-detected targets, and video-detected targets is confirmed as a common detection target, and the status information of the common detection target is obtained.

[0023] Furthermore, after obtaining the common detection target and its status information, it also includes:

[0024] The common detection target is identified as a newly entered state, and the status information of the common detection target is tracked in real time.

[0025] When the common detection target persists for several tracking state periods and the difference between any two of the first state information, second state information and third state information of the common detection target is less than a threshold, the common detection target is determined to be in a stable state.

[0026] When the jointly detected target does not exist continuously within a certain number of tracking state cycles and / or the difference between any two of the first state information, second state information, and third state information of the jointly detected target is greater than a threshold, the jointly detected target is determined to be in an exit state.

[0027] Furthermore, matching the target detected by the lidar and its second state information, the target detected by the millimeter-wave radar and its third state information, and the target detected by the video and its first state information with each other and associating them with the vehicle's predicted trajectory also includes:

[0028] When the matching of the target detected by the lidar and its second state information, the target detected by the millimeter-wave radar and its third state information, and the target detected by the video and its first state information fails, or when the association with the predicted trajectory of the vehicle fails, it is an invalid state. The invalid state indicates that no common detection target and its state information have been obtained.

[0029] Furthermore, before acquiring the lidar point cloud and the millimeter-wave radar point cloud, the method further includes the following steps:

[0030] Obtain the sampling frequency of the lidar and the sampling frequency of the millimeter-wave radar;

[0031] When the sampling frequency of the lidar is consistent with the sampling frequency of the millimeter-wave radar, the lidar point cloud and the millimeter-wave radar point cloud are acquired.

[0032] Furthermore, the vehicle emergency braking control strategy based on the commonly detected target and its state information includes:

[0033] When the state information of the common detection target meets the braking conditions, the vehicle's braking data is acquired;

[0034] If vehicle braking data is not obtained in the first instance, limit engine torque;

[0035] If the vehicle's braking data is not obtained within the second time period, automatic emergency braking will be performed.

[0036] Furthermore, the lidar point cloud and the millimeter-wave radar point cloud are used to acquire the velocity and distance of the jointly detected target; the velocity and distance of the jointly detected target include the longitudinal position, longitudinal velocity, longitudinal acceleration, lateral position, lateral velocity, and lateral acceleration of the jointly detected target; the lidar point cloud and the video data are used to acquire the target type, target height, target width, target depth, and azimuth of the jointly detected target.

[0037] The present invention also provides an electronic device, including a processor and a memory, wherein the processor is configured to perform the control method for automatic emergency braking of a vehicle as described in any of the preceding claims.

[0038] The present invention also provides a vehicle, comprising:

[0039] Camera, used to capture video data;

[0040] LiDAR, used to collect lidar point clouds;

[0041] Millimeter-wave radar is used to collect millimeter-wave radar point clouds;

[0042] An electronic device, including a processor and a memory, wherein the processor is configured to perform the control method for automatic emergency braking of a vehicle as described in any of the above embodiments;

[0043] A control terminal, used to execute instructions issued by the processor;

[0044] A CAN bus is used to connect one or more of the camera, the lidar, the millimeter-wave radar, and the control terminal to the processor.

[0045] The automatic emergency braking control method, electronic device, and vehicle provided by the present invention match the target detected by lidar and its second state information with the target detected by millimeter-wave radar and its third state information, as well as the target detected by video and its first state information, and associate them with the vehicle's predicted trajectory to obtain common target and state information. This improves the detection accuracy of the state information of the common target, avoids false positive and false negative detection results, and enhances the performance of automatic emergency braking. Attached Figure Description

[0046] The following drawings, which are incorporated herein by reference as part of this invention, are provided for understanding the invention. The drawings illustrate embodiments of the invention and their descriptions, serving to explain the principles of the invention.

[0047] In the attached image:

[0048] Figure 1 This is a schematic diagram of the structure of a vehicle according to an embodiment of the present invention;

[0049] Figure 2 A flowchart of a vehicle automatic emergency braking control method according to an embodiment of the present invention;

[0050] Figure 3 The information intersection of video data, lidar point cloud, and millimeter-wave radar point cloud according to an embodiment of the present invention is shown.

[0051] Figures 4A-4BA schematic diagram illustrating the matching of a target detected by lidar and its second state information, a target detected by millimeter-wave radar and its third state information, and a target detected by video and its first state information, and its association with the predicted trajectory of a vehicle, according to an embodiment of the present invention, is shown. Detailed Implementation

[0052] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention. However, it will be apparent to those skilled in the art that the invention can be practiced without these details. In other instances, certain technical features well-known in the art have not been described in order to avoid obscuring the invention.

[0053] It should be understood that the invention can be embodied in various forms and should not be construed as being limited to the embodiments set forth herein. Rather, providing these embodiments will make the disclosure thorough and complete, and will fully convey the scope of the invention to those skilled in the art. In the drawings, for clarity, the dimensions and relative dimensions of layers and regions may be exaggerated. The same reference numerals denote the same elements throughout.

[0054] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. When used herein, the singular forms “a,” “an,” and “the” are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the terms “comprising” and / or “including,” when used in this specification, identify the presence of the stated features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups. When used herein, the term “and / or” includes any and all combinations of the associated listed items.

[0055] To fully understand this invention, detailed steps and structures will be presented in the following description to illustrate the technical solution proposed by this invention. Preferred embodiments of the invention are described in detail below; however, in addition to these detailed descriptions, the invention may have other embodiments.

[0056] The 1R1V-based AEB scheme has several obvious drawbacks: First, because the parameters that the Radar and the camera excel at for target detection are different, false positives and false negatives are easily generated. Second, due to defects such as multipath reflection, the Radar is prone to producing many false targets or "ghosting" images, making it impossible for the AEB system to determine whether they are valid targets. Third, the camera's principle for detecting target distance and speed is generally to extract "corner points" from the image matrix of adjacent frames and perform feature matching, and then calculate the target's distance and speed based on visual SLAM or triangulation principles. When the relative speed between the vehicle equipped with the AEB system and the target vehicle is too high (e.g., greater than 80 kPH) or the ambient light is insufficient, the success rate of "corner point" matching in adjacent frames drops significantly, and the camera's accuracy in detecting target distance and speed will decrease. The higher the relative speed, the greater the decrease, and ultimately, it will be impossible to match the target detected by the Radar.

[0057] To address the above problems, the present invention provides a control method, electronic device, and vehicle for automatic emergency braking (AEB).

[0058] Below, first refer to Figure 1 A schematic diagram depicting the structure of a vehicle 100.

[0059] like Figure 1 As shown, vehicle 100 includes camera 110, lidar 120, millimeter-wave radar 130, electronic equipment 140, control terminal 150 and CAN bus 160.

[0060] In one embodiment, camera 110 is used to acquire video data; the type of camera 110 includes, but is not limited to, long-range cameras, mid-range cameras, stereo cameras, wide-angle cameras (e.g., fisheye cameras), surround cameras (e.g., 360-degree cameras), infrared cameras, or combinations thereof. LiDAR 120 is used to acquire lidar point clouds; the type of lidar 120 includes, but is not limited to, mechanical lidar, solid-state lidar, hybrid solid-state lidar, or combinations thereof. Millimeter-wave radar 130 is used to acquire millimeter-wave radar point clouds; the type of millimeter-wave radar 130 includes, but is not limited to, pulse radar, continuous-wave radar, pulse compression radar, frequency-agile radar, or combinations thereof.

[0061] In one embodiment, the electronic device 140 includes a processor 141 and a memory 142. The processor 141 can be implemented as software, hardware, firmware, or any combination thereof, and can use one or more application-specific integrated circuits (ASICs), one or more general-purpose integrated circuits, one or more microprocessors, one or more programmable logic devices, or any combination of the foregoing circuits and / or devices, or other suitable circuits or devices. Furthermore, the processor 141 can control other components in the vehicle 100 to perform corresponding steps of the methods described below. The memory 142 can be a flash memory card, solid-state memory, hard disk, etc. It can be volatile memory and / or non-volatile memory, removable memory and / or non-removable memory, etc.

[0062] In one embodiment, the CAN bus 160 is used to connect one or more of the camera 110, LiDAR 120, millimeter-wave radar 130, and control terminal 150 to the processor 141. Specifically, the camera 110 is connected to the processor 141 via LVDS, the LiDAR 120 is connected to the processor 141 via automotive Ethernet, the millimeter-wave radar 130 is connected to the processor 141 via the CAN bus 160, and the control terminal 150 is connected to the processor 141 via the CAN bus 160.

[0063] In one embodiment, the control terminal 150 includes, but is not limited to, an engine and / or a brake.

[0064] It should be understood that Figure 1 The components included in the vehicle 100 shown are merely illustrative and may include more or fewer components. This invention is not limited thereto.

[0065] To address the shortcomings of the 1R1V-based AEB scheme, this invention also provides a control method for automatic emergency braking of vehicles, such as... Figure 2 As shown, it includes:

[0066] Step S210: Acquire video data, lidar point cloud, and millimeter-wave radar point cloud collected at the same time;

[0067] Step S220: Based on the video data, obtain the video detection target and its first state information; based on the lidar point cloud, obtain the lidar detection target and its second state information; based on the millimeter-wave radar point cloud, obtain the millimeter-wave radar detection target and its third state information.

[0068] Step S230: Match the target detected by the lidar and its second state information, the target detected by the millimeter-wave radar and its third state information, and the target detected by the video and its first state information with each other and associate them with the vehicle's predicted trajectory to obtain common detection target and its state information;

[0069] Step S240: Execute the vehicle emergency braking control strategy based on the common detection target and its status information.

[0070] For example, before performing step S210, the method further includes the steps of: acquiring the sampling frequency of the lidar and the sampling frequency of the millimeter-wave radar; and acquiring the lidar point cloud and the millimeter-wave radar point cloud when the sampling frequency of the lidar and the sampling frequency of the millimeter-wave radar are consistent.

[0071] In one embodiment, the state variables of the millimeter-wave radar (Radar) and lidar (Lidar) tracks are adjusted from a previous time step to the current time step T to synchronize the sampling frequencies of the Radar and Lidar. The time step T can be used to calculate the target's velocity. and acceleration This necessitates a motion model to describe the changes of state variables over time. Both Radar and LiDAR trackers employ six state variables to describe the target's state—its longitudinal and lateral position, velocity, and acceleration. Both trackers use a constant acceleration motion model, where the evolution of the longitudinal and lateral states is independent, and the corresponding acceleration remains constant between the previous and current time steps. In other words, the current acceleration state equals the previous acceleration state, and the current velocity state equals the previous velocity plus the product of the previous acceleration and Delta time. Let the target's state be x. k The state equations at time k and time k-1 are:

[0072]

[0073] in, These are the target position, velocity, and acceleration observed by Radar and Lidar in the Cartesian coordinate system, respectively, and T is the time interval between time k and k-1.

[0074] Next, step S210 is executed: acquire video data, lidar point cloud, and millimeter-wave radar point cloud collected at the same time.

[0075] In one embodiment, processor 141 acquires video data collected by camera 110 via LVDS, acquires LiDAR point cloud collected by LiDAR 120 via vehicle Ethernet, and acquires millimeter-wave radar point cloud collected by millimeter-wave radar 130 via CAN bus 160.

[0076] Next, step S220 is executed: based on the video data, the video detection target and its first state information are obtained; based on the lidar point cloud, the lidar detection target and its second state information are obtained; and based on the millimeter-wave radar point cloud, the millimeter-wave radar detection target and its third state information are obtained.

[0077] In one embodiment, the video data is used to acquire the target type, target height, target width, target depth, and azimuth of the video detection target. Processing the video data yields one or more video detection targets and first state information for those targets, including camera state functions for the one or more video detection targets. in d represents the target azimuth, d represents the target depth / length, h represents the target height, w represents the target width, and t represents the target type.

[0078] In one embodiment, the lidar point cloud is used to acquire the velocity, range, target type, target height, target width, target depth, and azimuth of the lidar-detected target. Processing the lidar point cloud yields one or more lidar-detected targets and their second state information, which includes the LiDAR state functions of the one or more lidar-detected targets. as well as in, These represent the horizontal and vertical positions of the target, respectively. These are the target's horizontal and vertical velocities, respectively. These are the target's lateral and longitudinal accelerations, respectively. d represents the target azimuth, d represents the target depth / length, h represents the target height, w represents the target width, and t represents the target type.

[0079] In one embodiment, the millimeter-wave radar point cloud is used to acquire the velocity and range of the target detected by the millimeter-wave radar. Processing the millimeter-wave radar point cloud yields one or more millimeter-wave radar-detected targets, and third-state information of these targets. The third-state information includes the Radar state functions of the one or more millimeter-wave radar-detected targets. in, These represent the horizontal and vertical positions of the target, respectively. These are the target's horizontal and vertical velocities, respectively. These represent the target's lateral and longitudinal accelerations, respectively.

[0080] Therefore, as Figure 3 As shown, the lidar point cloud and the millimeter-wave radar point cloud are redundant to obtain the velocity and distance of the jointly detected target; the lidar point cloud and the video data are also redundant to obtain the target type, target height, target width, target depth, and azimuth of the jointly detected target. This avoids false positives and false negatives caused by the inconsistency of the parameters that the radar and the camera are good at for target detection, thus improving detection accuracy and enhancing the performance of AEB.

[0081] Next, step S230 is executed: the target detected by the lidar and its second state information, the target detected by the millimeter-wave radar and its third state information, and the target detected by the video and its first state information are matched with each other and associated with the vehicle's predicted trajectory to obtain common detected targets and their state information. Here, the vehicle's predicted trajectory refers to the path planned by the vehicle's intelligent driving system for the upcoming journey.

[0082] For example, the one or more lidar-detected targets and their second state information are matched with the one or more millimeter-wave radar-detected targets and their third state information to obtain a pair of lidar-detected targets and millimeter-wave radar-detected targets, wherein the pair of lidar-detected targets and millimeter-wave radar-detected targets are the same object.

[0083] In one embodiment, the Radar and LiDAR output their respective detected targets and their state information, along with an active, real-time updated prediction parameter. The real-time updated prediction parameter essentially compares the target state directly detected by the sensors at the current moment with the target state predicted at the previous moment to verify the reasonableness of the previous prediction. If unreasonable, it corrects relevant prediction parameters, such as the Kalman gain, and also predicts the target state at the next moment. Target association aims to correlate the target detection results of the Radar and LiDAR with the vehicle's predicted trajectory.

[0084] To perform correlation matching for targets detected by LiDAR and Radar, in order to effectively correlate the targets detected by Radar and LiDAR with the vehicle's predicted trajectory, the detected targets must pass through a threshold. This threshold is a multi-dimensional space. Since Radar and LiDAR have relatively high accuracy in detecting target speed and distance, these parameters are used as dimensions in the multi-dimensional space, including the horizontal and vertical position, speed, and acceleration dimensions of the detected targets.

[0085] Wherein, the Lidar state function is The Radar state function is Define the distance between the Lidar and Radar state functions as D. (1) =L (1) -R, when When D (1) When the minimum value is obtained, the target detected by Radar is associated with the target detected by LiDAR, and the same ID is assigned to them in the AEB system. L-R =ID L =ID R .

[0086] For example, the one or more lidar-detected targets and their second state information are matched with the one or more video-detected targets and their first state information to obtain a pair of lidar-detected targets and video-detected targets, wherein the pair of lidar-detected targets and video-detected targets are the same object.

[0087] The targets detected by the LiDAR and camera also need to be associated and matched. In order to effectively associate the targets detected by the camera and LiDAR with the predicted trajectory, the detected targets must pass through a threshold. This threshold is a multi-dimensional space. Since the camera and LiDAR have high accuracy in detecting the target azimuth, depth, height, width and target type, these parameters are used as dimensions in the multi-dimensional space, including the horizontal and vertical position, velocity and acceleration dimensions of the detected targets.

[0088] The Lidar state function is The camera state function is Define the distance between the LiDAR and the camera state function as D. (2) =L (2) -C, when When D (2) When the minimum value is obtained, the target detected by the camera is associated with the target detected by the LiDAR, and they are assigned the same ID. L-C =ID L =ID C .

[0089] After associating the Radar with the LiDAR, and then the LiDAR with the camera, for the same target, there will be an ID. T =ID L-C =ID L-R .

[0090] For example, a set of LiDAR-detected targets, millimeter-wave radar-detected targets, and video-detected targets are obtained based on the pair of LiDAR-detected targets and millimeter-wave radar-detected targets and the pair of LiDAR-detected targets and video-detected targets; the set of LiDAR-detected targets, millimeter-wave radar-detected targets, and video-detected targets are respectively associated with the vehicle's predicted trajectory to obtain common detection targets and their state information.

[0091] For example, associating the set of lidar-detected targets, millimeter-wave radar-detected targets, and video-detected targets with the vehicle's predicted trajectory to obtain common detection targets and their status information includes: obtaining a first distance between the lidar-detected target and the vehicle's predicted trajectory, obtaining a second distance between the millimeter-wave radar-detected target and the vehicle's predicted trajectory, and obtaining a third distance between the video-detected target and the vehicle's predicted trajectory; when the difference between any two of the first distance, the second distance, and the third distance is less than a threshold, confirming that the set of lidar-detected targets, millimeter-wave radar-detected targets, and video-detected targets are common detection targets, and obtaining the status information of the common detection targets.

[0092] In one embodiment, Figure 4A This describes the target positions detected by each sensor and the predicted vehicle trajectory. The association between the targets detected by Radar and LiDAR and the predicted vehicle trajectory is achieved through an association matrix. This matrix records the distance between the target detected by each sensor and the predicted vehicle trajectory at the current time, such as... Figure 4B As shown. Then, the algorithm searches the correlation matrix for the item with the smallest distance between the vehicle's predicted trajectory and the target detected by the sensor. In this embodiment, the distance between the Radar-detected target and the vehicle's predicted trajectory is D = 0.12, the distance between the Lidar-detected target and the vehicle's predicted trajectory is D = 0.11, and the distance between the camera-detected target and the vehicle's predicted trajectory is D = 0.13. The Lidar-detected target and the Radar-detected target, as well as the camera-detected target and the vehicle's predicted trajectory, in this region are successfully matched and determined by the correlation algorithm to be the same target. Before matching, the coordinates of the Lidar and Radar need to be transformed to the same vehicle coordinate system. In subsequent algorithms, this vehicle's predicted trajectory and the corresponding detected target will be tracked synchronously.

[0093] For example, after obtaining the common detection target and its status information, the method further includes: determining the common detection target as a newly entered state and tracking the status information of the common detection target in real time; when the common detection target exists continuously for several tracking state periods and the difference between any two of the first, second, and third status information of the common detection target is less than a threshold, determining the common detection target as a stable state; when the common detection target does not exist continuously for several tracking state periods and / or the difference between any two of the first, second, and third status information of the common detection target is greater than a threshold, determining the common detection target as an exited state.

[0094] For example, matching the LiDAR-detected target and its second state information, the millimeter-wave radar-detected target and its third state information, and the video-detected target and its first state information with each other and associating them with the vehicle's predicted trajectory further includes: when the LiDAR-detected target and its second state information, the millimeter-wave radar-detected target and its third state information, and the video-detected target and its first state information fail to match, and the association with the vehicle's predicted trajectory fails, it is an invalid state, which indicates that no common detection target and its state information have been obtained.

[0095] Next, following step S230, the method further includes: adding derived signals, wherein the derived signals include one or more of the following: road curvature estimation based on vehicle body data and data results from the previous step; conversion between the vehicle coordinate system (VCS) coordinate system and polar or Cartesian coordinate systems; target confidence; target presence probability; detection accuracy; or target priority level. The derived signals are primarily used to provide vehicle status information to improve the target perception capabilities of cameras, LiDAR, and Radar, and to eliminate sensor perception errors.

[0096] Next, step S240 is executed: the vehicle emergency braking control strategy is executed based on the common detection target and its status information.

[0097] For example, the vehicle emergency braking control strategy based on the common detection target and its status information includes: acquiring vehicle braking data when the status information of the common detection target meets the braking conditions; limiting engine torque when vehicle braking data is not acquired in a first time period; and performing automatic emergency braking when vehicle braking data is still not acquired in a second time period.

[0098] In one embodiment, based on the target perception results from the camera, LiDAR, and Radar in the above steps, AEB can output the status of highly reliable targets, i.e., which targets require braking action, which targets need to be tracked in real time, and which targets can be discontinued. Based on this step, target and trajectory are tracked at a functional level. When the target meets the braking conditions, if the driver does not take braking avoidance measures, the AEB system limits engine torque. If the driver still does not take measures, the AEB system will use the braking system to perform emergency braking, ensuring the vehicle stops safely to avoid a collision with a certain braking deceleration. The deceleration value is based on information such as the current speed of the main vehicle, the time to collision (TTC) between the main vehicle and the target vehicle, the current speed of the main vehicle, and the braking distance under the operating conditions.

[0099] The automatic emergency braking control method for vehicles provided by the present invention matches the target detected by lidar and its second state information with the target detected by millimeter-wave radar and its third state information, as well as the target detected by video and its first state information, and associates them with the vehicle's predicted trajectory to obtain common detection target and its state information. This improves the detection accuracy of the state information of the common detection target, avoids false positive and false negative detection results, and enhances the performance of automatic emergency braking.

[0100] The present invention has been described through the above embodiments. However, it should be understood that the above embodiments are for illustrative purposes only and are not intended to limit the invention to the scope of the described embodiments. Furthermore, those skilled in the art will understand that the present invention is not limited to the above embodiments, and many more variations and modifications can be made based on the teachings of the present invention, all of which fall within the scope of protection claimed by the present invention. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A control method for automatic emergency braking of a vehicle, characterized in that, include: Acquire video data, lidar point cloud, and millimeter-wave radar point cloud collected at the same time; Based on the video data, one or more video detection targets and their first state information are obtained; based on the lidar point cloud, one or more lidar detection targets and their second state information are obtained; based on the millimeter-wave radar point cloud, one or more millimeter-wave radar detection targets and their third state information are obtained. The one or more lidar-detected targets and their second state information are matched with the one or more millimeter-wave radar-detected targets and their third state information to obtain a pair of lidar-detected targets and millimeter-wave radar-detected targets, wherein the pair of lidar-detected targets and millimeter-wave radar-detected targets are the same object. The one or more lidar-detected targets and their second state information are matched with the one or more video-detected targets and their first state information to obtain a pair of lidar-detected targets and video-detected targets, wherein the pair of lidar-detected targets and video-detected targets are the same object; Based on the pair of lidar-detected targets and millimeter-wave radar-detected targets, and the pair of lidar-detected targets and video-detected targets, a set of lidar-detected targets, millimeter-wave radar-detected targets, and video-detected targets are obtained; The set of lidar-detected targets, millimeter-wave radar-detected targets, and video-detected targets are respectively associated with the vehicle's predicted trajectory to obtain common detected targets and their status information; The vehicle emergency braking control strategy is executed based on the common detection target and its status information.

2. The control method as described in claim 1, characterized in that, Associating the set of lidar-detected targets, millimeter-wave radar-detected targets, and video-detected targets with the vehicle's predicted trajectory to obtain common detected targets and their status information includes: Obtain a first distance between the target detected by the lidar and the predicted trajectory of the vehicle; obtain a second distance between the target detected by the millimeter-wave radar and the predicted trajectory of the vehicle; obtain a third distance between the target detected by the video and the predicted trajectory of the vehicle. When the difference between any two of the first distance, the second distance, and the third distance is less than a threshold, the group of LiDAR-detected targets, millimeter-wave radar-detected targets, and video-detected targets is confirmed as a common detection target, and the status information of the common detection target is obtained.

3. The control method as described in claim 1, characterized in that, After obtaining the common detection target and its status information, the following is also included: The common detection target is identified as a newly entered state, and the status information of the common detection target is tracked in real time. When the common detection target continues to exist within a number of tracking state cycles and the difference between any two of the first state information, second state information, and third state information of the common detection target is less than a threshold, the common detection target is determined to be in a stable state. When the jointly detected target does not exist continuously within a certain number of tracking state cycles and / or the difference between any two of the first state information, second state information, and third state information of the jointly detected target is greater than a threshold, the jointly detected target is determined to be in an exit state.

4. The control method as described in claim 1, characterized in that, The process of matching the target detected by the lidar and its second state information, the target detected by the millimeter-wave radar and its third state information, and the target detected by the video and its first state information with the vehicle's predicted trajectory also includes: When the matching of the target detected by the lidar and its second state information, the target detected by the millimeter-wave radar and its third state information, and the target detected by the video and its first state information fails, or when the association with the predicted trajectory of the vehicle fails, it is an invalid state. The invalid state indicates that no common detection target and its state information have been obtained.

5. The control method as described in claim 1, characterized in that, Before acquiring the lidar point cloud and the millimeter-wave radar point cloud, the following steps are also included: Obtain the sampling frequency of the lidar and the sampling frequency of the millimeter-wave radar; When the sampling frequency of the lidar is consistent with the sampling frequency of the millimeter-wave radar, the lidar point cloud and the millimeter-wave radar point cloud are acquired.

6. The control method as described in claim 1, characterized in that, The vehicle emergency braking control strategy based on the common detection target and its status information includes: When the state information of the common detection target meets the braking conditions, the vehicle's braking data is acquired; If vehicle braking data is not obtained in the first instance, limit engine torque; If the vehicle's braking data is not obtained within the second time period, automatic emergency braking will be performed.

7. The control method as described in claim 1, characterized in that, The lidar point cloud and the millimeter-wave radar point cloud are used to acquire the velocity and distance of the jointly detected target. The velocity and distance of the jointly detected target include the longitudinal position, longitudinal velocity, longitudinal acceleration, lateral position, lateral velocity, and lateral acceleration of the jointly detected target. The lidar point cloud and the video data are used to acquire the target type, target height, target width, target depth, and azimuth of the jointly detected target.

8. An electronic device, characterized in that, It includes a processor and a memory, wherein the processor is used to execute the control method for automatic emergency braking of a vehicle as described in any one of claims 1-7.

9. A vehicle, characterized in that, include: Camera, used to capture video data; LiDAR, used to collect lidar point clouds; Millimeter-wave radar is used to collect millimeter-wave radar point clouds; An electronic device, comprising a processor and a memory, wherein the processor is configured to perform the control method for automatic emergency braking of a vehicle as described in any one of claims 1-7; A control terminal, used to execute instructions issued by the processor; A CAN bus is used to connect one or more of the camera, the lidar, the millimeter-wave radar, and the control terminal to the processor.

Images