Automatic Emergency Breaking System for Vehicles

Abstract

Technology regarding automatic emergency braking (AEB) of an automobile is disclosed. The automatic emergency braking system outputs an automatic emergency braking signal by reflecting the slope of the road ahead. The road slope can be calculated based on changes in the elevation of the radar point cloud corresponding to the road boundary. The braking distance is calculated using vehicle speed information, road surface condition information, and slope information, and the collision risk can be evaluated using this value and the Time to Collision (TTC).

Description

Technology Field

[0001] Technology is disclosed regarding an automobile braking system, particularly an Automatic Emergency Braking (AEB) system that predicts a risk of collision and automatically brakes independently of the driver. Background Technology

[0002] The Automatic Emergency Braking (AEB) system assesses the risk of collision with objects ahead, such as other vehicles or pedestrians, during driving conditions. Based on this assessment, it sequentially provides warnings to the driver and automatically performs partial and full braking. Through this, the AEB system prevents collisions in situations where longitudinal collisions are at risk due to driver error or insufficient reaction time, and even if a collision is unavoidable, it reduces the collision speed to mitigate damage.

[0003] Korean Published Patent Application No. 10-2023-0102465, disclosed on July 7, 2023, discloses an automatic emergency braking technology that takes into account road surface conditions. The disclosed technology detects the relative distance from a vehicle to a preceding vehicle, detects the speed of the vehicle itself, acquires road surface condition information, compensates for the detected relative distance accordingly to calculate the Time to Collision (TTS), and performs automatic emergency braking according to the degree of collision risk.

[0004] However, in addition to road surface conditions caused by rain, the gradient of the road can significantly affect the time to collision because braking distances increase on downhill slopes. The problem to be solved

[0005] The proposed invention aims to operate an automatic emergency braking device by reflecting the road gradient.

[0006] Furthermore, the proposed invention aims to acquire road gradient and additional information necessary for automatic emergency braking by adding a simple sensor. means of solving the problem

[0007] According to one aspect of the proposed invention, an automatic emergency braking signal is output by reflecting the slope of the road ahead.

[0008] Depending on additional aspects, the slope of the road can be calculated based on changes in the elevation of the radar point cloud corresponding to the road boundary.

[0009] According to an additional aspect of the invention, the relative distance and relative speed to the vehicle ahead can be measured from the radar point cloud and reflected in the assessment of collision risk.

[0010] According to one aspect, braking distance can be calculated using vehicle speed information, road surface condition information, and gradient information, and collision risk can be evaluated using this value and Time to Collision (TTC). Furthermore, collision risk can be evaluated by reflecting vehicle speed and road surface condition information in addition to the road gradient. Effects of the invention

[0011] According to the proposed invention, an automatic emergency braking signal is output by reflecting the slope of the road ahead, thereby reflecting the risk associated with the braking distance shortened by the road slope. The slope of the road is calculated in real time based on changes in the elevation of the radar point cloud corresponding to the road boundary, and a safer vehicle can be realized by adding a 4D radar that is relatively inexpensive and can provide other complex information. Brief explanation of the drawing

[0012] FIG. 1 is a diagram illustrating an exemplary braking situation that serves as the background of the proposed invention. FIG. 2 is a block diagram illustrating the configuration of an automatic emergency braking device of a vehicle according to one embodiment. FIG. 3 is a flowchart illustrating the configuration of an embodiment of an automatic emergency braking control method for a vehicle implemented by one or more programs stored in memory (300). FIG. 4 is a flowchart illustrating the configuration of an inclination information acquisition operation according to one embodiment. FIG. 5 is a flowchart illustrating the configuration of a method for calculating relative distance and relative speed according to one embodiment. FIG. 6 is a block diagram illustrating the configuration of an automatic emergency braking device of a vehicle according to another embodiment. FIG. 7 is a flowchart illustrating the configuration of a road surface condition information acquisition step according to one embodiment. FIG. 8 is a flowchart illustrating the configuration of a collision risk calculation operation according to one embodiment. Specific details for implementing the invention

[0013] The foregoing and additional aspects are embodied through embodiments described with reference to the attached drawings. It is understood that the components of each embodiment may be combined in various ways within the embodiment or with components of other embodiments, unless otherwise stated or contradicted. Based on the principle that the inventor can appropriately define the concepts of terms to best describe his invention, the terms used in this specification and claims shall be interpreted in a meaning and concept consistent with the description or proposed technical idea.

[0014] Blocks referred to as 'circuits' or 'parts' with reference to block diagrams in this specification may be composed of hardware, such as dedicated semiconductors, gate arrays, or FPGAs, or parts thereof. One or more blocks may be implemented as a single piece of hardware. As another example, these blocks may be implemented in software as an information processing device in which a computation element executes program instructions stored in a memory element. Multiple blocks may be implemented as part of a program executed on the same computation element. As yet another example, these blocks may be implemented in a hybrid form in which part of the individual circuit is hardware and part is software. Furthermore, in a software implementation, the computation element may include digital signal processors, dedicated computation processors, artificial intelligence processing engines, dedicated artificial intelligence processors, or graphics processors, or a combination thereof to the extent possible.

[0016] Preferred embodiments of the present invention will be described in detail below with reference to the attached drawings.

[0017] <Explanation of Fig. 1>

[0018] FIG. 1 is a diagram illustrating an exemplary braking situation that serves as the background of the proposed invention. In the diagram, a stereoscopic imaging sensor, in this embodiment, a 4D radar (13) is mounted on the front of the vehicle (10). The 4D radar is also called an image radar and is a sensor that outputs a point cloud containing 3D position and additional Doppler information related to speed through a MIMO (Multiple-input-multiple-output) antenna array. The radar sensor periodically outputs point cloud frames. A road boundary can be detected from the point cloud frame output by the radar at position A. At this time, the elevation of the detected road boundary can be stored. As the vehicle travels, a road boundary can be detected from the point cloud frame output by the radar when it reaches position B. At this time, the elevation of the detected road boundary becomes smaller than the previously stored value. The proposed invention can calculate the slope of the road ahead from the change in the elevation value of the road boundary of the road ahead detected by the stereoscopic imaging sensor.

[0019] <1>

[0020] FIG. 2 is a block diagram illustrating the configuration of an automatic emergency braking device for a vehicle according to one embodiment. The illustrated automatic emergency braking device may be configured as a single module in the vehicle. In one embodiment, the automatic emergency braking device module may be connected to a vehicle network, such as a Car Area Network (CAN) or Vehicle Ethernet, via a cable harness, in a form in which a circuit board with an intelligent processor, such as a sensor and a microprocessor, is fixed inside a housing. As illustrated, the automatic emergency braking device for a vehicle according to one embodiment may include a stereoscopic imaging sensor (500), one or more processors (100), and a memory (300).

[0021] The stereoscopic imaging sensor (500) periodically outputs a point cloud of the road ahead. For example, the stereoscopic imaging sensor may be a depth camera. As another example, the stereoscopic imaging sensor may be a Lidar sensor. In the illustrated embodiment, the stereoscopic imaging sensor is a 4D radar. A 4D radar, also known as an image radar, is a sensor that outputs a point cloud frame containing 3D position and additionally Doppler information related to velocity through a MIMO (Multiple-input-multiple-output) antenna array. One or more processors (100) may include, for example, a multicore processor, an AI-dedicated processor such as a Neural Processing Unit (NPU), and a digital signal processor. One or more of these may be connected to the stereoscopic imaging sensor to receive and process data output from it. For example, the AI ​​processor may detect the boundaries of the road ahead from the point cloud frame data output from the stereoscopic imaging sensor. For example, a multicore processor can calculate the slope of the road ahead from a sequence of point clouds corresponding to the road ahead detected by an artificial intelligence processor. For example, a digital signal processor can generate a radar point cloud by processing radar signals output from the receiving antennas of a 4D radar.

[0022] Memory (300) can store one or more programs configured to be executed by one or more processors, and associated data. In one embodiment, memory (300) may be designed in a hierarchical structure including non-volatile memory such as flash memory and volatile memory such as SRAM (Static Random Access Memory), and may be designed in a structure including local memory connected to each processor and main memory shared by all processors. Programs stored in memory may be modularized by one or more functions. They may exist in spatially different memories. That is, modules stored in local memory connected to each processor and modules stored in main memory may cooperate with each other. One or more processors may evaluate the risk of a forward collision and output an automatic emergency braking signal. For example, the automatic emergency braking signal may be output to a direct braking device.

[0023] Optionally, an automatic emergency braking device according to one embodiment may further include a communication circuit (700). The communication circuit (700) may include a modem that can be connected to a vehicle network, such as a CAN (Car Area Network) or vehicle Ethernet. In this case, an automatic emergency braking signal output by one or more processors may be output to the communication circuit (700) and transmitted to the braking device through the vehicle network.

[0024] FIG. 3 is a flowchart illustrating the configuration of an embodiment of an automatic emergency braking control method for a vehicle implemented by one or more programs stored in memory (300). In one aspect, an automatic emergency braking signal is output by reflecting the slope of the road ahead. In this aspect, the automatic emergency braking control method for a vehicle according to one embodiment includes a slope information acquisition operation (S100), a collision risk assessment operation (S800, S950), and an emergency braking signal generation operation (S970). These operations may be performed by independent program modules or by functional modules mixed within a single program.

[0025] First, a point cloud frame is obtained from a stereoscopic imaging sensor, and in the illustrated embodiment, from a 4D radar (S910). However, the proposed invention is not limited to a 4D radar, and a point cloud can be obtained from various stereoscopic imaging sensors as described above.

[0026] In the slope information acquisition operation (S100), one or more processors, or multiple processors, acquire slope information of the road ahead from a point cloud. For example, slope information can be acquired from a LiDAR image capturing the road ahead. A road sidewall is detected in a LiDAR Depth Image, and slope information can be acquired from changes in the elevation of the road sidewall in consecutive LiDAR image frames.

[0027] Depending on additional aspects, the slope of the road can be calculated based on changes in elevation of the radar point cloud corresponding to the road boundary. For example, slope information can be obtained from the 4D radar point cloud of the road ahead. The 4D radar point cloud contains Doppler information and radar cross-section (RCS) information, which can increase reliability in object detection, in this case, road boundary detection. In addition, since it contains Doppler information related to relative speed, not only the relative distance but also the relative speed with the vehicle directly ahead can be calculated from this point cloud. Furthermore, compared to other sensors such as cameras or lidar, radar has higher reliability of point cloud data at long distances, which can increase reliability when applied to the field of automatic emergency braking. The road sidewall is detected in the 4D radar point cloud, and slope information can be obtained from changes in elevation of the road sidewall in consecutive point cloud frames.

[0028] Subsequently, one or more processors, or multiple processors, calculate a collision risk by reflecting the slope information obtained in the collision risk calculation operation (S800) and evaluate it (S950). In the emergency braking signal generation operation (S970), one or more processors, or multiple processors, output an emergency braking signal according to the evaluated collision risk.

[0029] <Fig. 4>

[0030] FIG. 4 is a flowchart illustrating the configuration of a slope information acquisition operation according to an embodiment applied to an additional aspect. As illustrated, the slope information acquisition operation (S100) according to an embodiment may include a road boundary detection operation (S110), an elevation calculation operation (S130), and an elevation difference calculation operation (S150).

[0031] In the road boundary detection operation (S110), one or more processors detect road boundaries from the radar point cloud. In one embodiment, an artificial intelligence processor may be trained and applied to detect road boundaries from the radar point cloud. In particular, the part where a horizontal road meets a vertical wall in an underpass, or guardrails installed along the side of a road in a general road, react sensitively to radar signals and are well reflected in the point cloud.

[0032] In the elevation calculation operation (S150), one or more processors calculate the elevation of the detected road boundary. In one embodiment, a multicore processor may obtain an elevation value by averaging the elevation values ​​of samples among the point clouds belonging to the road boundary output by the artificial intelligence processor that have an RCS value of at least a certain level, and store the result in memory.

[0033] In the elevation difference calculation operation (S170), one or more processors generate slope information of the road ahead from the difference values ​​of elevation values ​​of consecutive radar point cloud frames. In one embodiment, a multicore processor may determine the difference between consecutive frames of elevation values ​​calculated for each radar point cloud frame as the slope value.

[0034] <5>

[0035] In the embodiment illustrated in FIG. 3, the automatic emergency braking device determines whether to brake based on the relative distance and relative speed with respect to the vehicle ahead. According to an additional aspect of the invention, the relative distance and relative speed with respect to the vehicle ahead can be measured from a radar point cloud and reflected in the assessment of collision risk. FIG. 5 is a flowchart illustrating the configuration of a method for calculating relative distance and relative speed according to an embodiment reflecting this aspect. As illustrated, the method for calculating relative distance and relative speed according to an embodiment includes a vehicle ahead detection step (S710) and a relative speed and relative distance calculation step (S730).

[0036] In FIG. 3, one or more processors, in the illustrated embodiment, an artificial intelligence processor detects a vehicle ahead from the radar point cloud frame obtained in step S910 (710). The technology of classifying and detecting objects as one of vehicles, bicycles, people, or animals is now a well-known technology in the field of autonomous driving. Subsequently, relative speed and relative distance are calculated from the radar point cloud of the vehicle and reflected in the collision risk assessment (S730). Since the radar point cloud includes line-of-sight speed and line-of-sight distance values, relative speed and relative distance are obtained from them, and relative acceleration values ​​can also be obtained from the change value of relative speed on the time axis. This process can be repeated for all vehicles detected ahead (S750). Subsequently, the value with respect to the vehicle's direction of travel can be determined as a representative value (S770).

[0037] <Fig. 6>

[0038] Depending on additional aspects, collision risk can be evaluated by reflecting road surface condition information in addition to the road gradient. Furthermore, collision risk can be evaluated by reflecting vehicle speed and road surface condition information in addition to the road gradient.

[0039] FIG. 6 is a block diagram illustrating the configuration of an automatic emergency braking device for a vehicle according to another embodiment. In the embodiment illustrated in comparison with the embodiment of FIG. 2, one or more processors include a multi-core processor (110), an artificial intelligence processor (130), and a digital signal processor (150). Also, the memory (300) includes a flash memory (310) and an SRAM (330). The components of the device process data input and output through a system bus (900). The communication circuit (700) may include a modem that can be connected to a vehicle network, such as a CAN (Car Area Network) or vehicle Ethernet.

[0040] <Fig. 7>

[0041] According to an additional aspect, the automatic emergency braking device according to the illustrated embodiment further includes a camera sensor (510). The camera sensor (510) acquires a high-resolution image of the front and may be composed of a plurality of cameras having different angles of view.

[0042] Referring again to FIG. 3, vehicle speed information can be obtained in step S300. Vehicle speed information can be received, for example, from another component of the vehicle, such as a speedometer, by the multicore processor (110) in FIG. 6 connecting to the vehicle network through the communication circuit (700).

[0043] Additionally, road surface condition information may be obtained in step S500. FIG. 7 is a flowchart illustrating the configuration of the road surface condition information acquisition step according to one embodiment. The embodiment of FIG. 7 may be implemented based on the device of FIG. 6. As illustrated, a camera image frame may be obtained from the camera (510) of FIG. 6 (S510). Subsequently, for example, an artificial intelligence processor (130) may detect the road surface in the camera image frame (S530). Subsequently, the artificial intelligence processor (130) determines the condition of the detected road surface portion (S550). The determination of the road surface condition may also be determined by the artificial intelligence processor as a classification problem for the detected road surface image.

[0044] Returning to FIG. 3, in the collision risk assessment operation (S800, S590), for example, a multicore processor assesses the collision risk by reflecting the acquired slope information. Generally, the collision risk can be assessed by the Time To Collision (TTS), which is the estimated time until a collision with the vehicle ahead. The Time To Collision is determined by reflecting the relative distance, relative speed, and acceleration relative to the vehicle ahead. In one embodiment, the Time To Collision value may be adjusted to reflect the slope. For example, in the case of a downhill slope, the Time To Collision value may be reduced in proportion to the slope.

[0045] Afterwards, the time until collision is compared with a reference value (S950). If the time until collision is greater than or equal to the reference value, the process returns to the beginning, otherwise, an emergency braking signal is output (S970).

[0046] As another example, instead of adjusting the time to collision, the threshold at stage S950 can be adjusted. For instance, the threshold can be reduced as the slope becomes more downhill.

[0047] <Fig. 8>

[0048] According to one aspect, the braking distance is calculated using vehicle speed information, road surface condition information, and slope information, and the collision risk can be evaluated using this value and the time to collision (TTC). FIG. 8 is a flowchart illustrating the configuration of a collision risk calculation operation according to one embodiment. As illustrated, the collision risk calculation operation according to one embodiment first obtains the relative distance and relative speed to the vehicle ahead from a point cloud within the range of the detected vehicle ahead (S810). This information can be immediately obtained from the line-of-sight distance information and Doppler information included in the radar point cloud information output by the 4D radar signal. Subsequently, the time to collision (TTS) for the vehicle ahead is calculated from these values ​​(S830). In this embodiment, the collision risk can be defined as the reciprocal of the time to collision value. Subsequently, the time to collision is corrected by reflecting the slope (S850). In this embodiment, the reference value is defined as a function of the braking distance; for example, the reference value can be the braking distance divided by the vehicle speed. This must be reflected during acceleration motion. Subsequently, the braking distance is adjusted based on road conditions and vehicle speed, and the threshold value serving as the standard for collision risk assessment is adjusted accordingly. For example, if the road surface is wet during rain, the braking distance shortens; therefore, the threshold value must be increased even if the time to collision remains the same.

[0049] In the illustrated embodiment, the slope is reflected in the time to collision (TTS), but it may also be reflected in the braking distance. If the slope is downhill, the frictional force is reduced by the force component in the direction of the ground, and consequently, the braking distance may be increased. This is calculated and reflected in the braking distance, and the reference value may be adjusted accordingly.

[0050] <eos>

[0051] Although the present invention has been described above with reference to embodiments with reference to the accompanying drawings, it is not limited thereto and should be interpreted to encompass various variations that can be obviously derived from them by those skilled in the art. The claims are intended to encompass such variations. Explanation of the symbols

[0052] 10 : Vehicle 13: Stereoscopic imaging sensor 100: One or more processors 110: Multicore processor 130 : Artificial Intelligence Processor 150 : Digital signal processor 300 : Memory 310: Flash memory 330 : SRAM 500: Stereo imaging sensor 510 : Camera 700: Communication circuit< / eos>

Claims

Claim 1 An automatic emergency braking device for a vehicle comprising: a stereoscopic imaging sensor for acquiring a point cloud of a road ahead; one or more processors connected to the stereoscopic imaging sensor; and a memory for storing one or more programs configured to be executed by said one or more processors, wherein said one or more programs include: a slope information acquisition module for acquiring slope information of a road ahead from a point cloud; a collision risk assessment module for evaluating a collision risk by reflecting the acquired slope information; and an emergency braking signal generation module for outputting an emergency braking signal according to the evaluated collision risk. Claim 2 The automatic emergency braking device of a vehicle, wherein the stereoscopic imaging sensor is: a four-dimensional radar that outputs a radar point cloud. Claim 3 The automatic emergency braking device of a vehicle according to claim 2, wherein the slope information acquisition module calculates the slope of the road based on a change in elevation of a radar point cloud corresponding to the road boundary. Claim 4 An automatic emergency braking device for a vehicle according to claim 3, wherein the slope information acquisition module comprises: a road boundary detection module that detects a road boundary from a radar point cloud; an altitude calculation module that calculates the altitude of the detected road boundary; and an altitude difference calculation module that generates slope information of a road ahead from the difference values ​​of altitude values ​​of consecutive radar point cloud frames. Claim 5 An automatic emergency braking device for a vehicle according to claim 2, wherein one or more programs further include a relative distance and relative speed calculation module that detects a vehicle ahead from a radar point cloud and calculates a relative distance and relative speed with respect to the vehicle ahead from the point clouds of the detected vehicle ahead and reflects this in a collision risk assessment. Claim 6 The automatic emergency braking device according to claim 5, wherein the device further comprises a camera that photographs the front of the vehicle; and the one or more programs further comprise a road surface condition information acquisition module that acquires road surface condition information from an image captured by the camera and reflects it in a collision risk assessment. Claim 7 In claim 6, the collision risk evaluation module evaluates the collision risk by reflecting the relative distance and relative speed with respect to a vehicle ahead, vehicle speed information and road surface condition information, and additionally, the slope information of the road ahead. Claim 8 An automatic emergency braking device for a vehicle according to claim 7, wherein the device further comprises: a communication circuit connected to one of the vehicle internal control modules via a vehicle network, and further comprises a speed information acquisition module in which one or more programs acquire speed information of the vehicle via the communication circuit. Claim 9 An automatic emergency braking device of a vehicle according to claim 5, wherein the collision risk evaluation module calculates the braking distance using vehicle speed information, road surface condition information, and slope information, and evaluates the collision risk using this value and the time to collision (TTC). Claim 10 An automatic emergency braking control method for a vehicle, implemented in one or more programs of an automatic emergency braking device comprising: a stereoscopic imaging sensor for acquiring a point cloud of a road ahead; one or more processors connected to the stereoscopic imaging sensor; and a memory for storing one or more programs configured to be executed by said one or more processors, wherein the method comprises: an inclination information acquisition operation for acquiring inclination information of a road ahead from a point cloud; a collision risk assessment operation for evaluating a collision risk by reflecting the acquired inclination information; and an emergency braking signal generation operation for outputting an emergency braking signal according to the evaluated collision risk. Claim 11 In claim 10, the stereoscopic imaging sensor is: a four-dimensional radar that outputs a radar point cloud, an automatic emergency braking control method for a vehicle. Claim 12 In claim 11, the slope information acquisition module calculates the slope of the road based on the change in elevation of a radar point cloud corresponding to the road boundary. Automatic emergency braking control method for a vehicle. Claim 13 The method of claim 12, wherein the slope information acquisition operation comprises: a road boundary detection operation for detecting a road boundary from a radar point cloud; an altitude calculation operation for calculating the altitude of the detected road boundary; and an altitude difference calculation operation for generating slope information of a road ahead from the difference values ​​of altitude values ​​of consecutive radar point cloud frames. Claim 14 The automatic emergency braking control method of a vehicle according to claim 11, wherein the method implemented in one or more programs further comprises: a relative distance and relative speed calculation operation for detecting a vehicle ahead from a radar point cloud and calculating a relative distance and relative speed with respect to the vehicle ahead from the point clouds of the detected vehicle ahead and reflecting this in a collision risk assessment. Claim 15 The automatic emergency braking control method according to claim 14, wherein the device further comprises a camera that photographs the front of the vehicle; and the method implemented in one or more programs further comprises a road surface condition information acquisition operation that acquires road surface condition information from an image captured by the camera and reflects it in a collision risk assessment. Claim 16 In claim 15, the collision risk assessment operation evaluates the collision risk by reflecting the relative distance and relative speed with respect to a vehicle ahead, vehicle speed information and road surface condition information, and additionally, the slope information of the road ahead. Claim 17 The automatic emergency braking control method of a vehicle according to claim 16, wherein the device further comprises a communication circuit that connects to one of the vehicle internal control operations through a vehicle network, and the method implemented in one or more programs further comprises a speed information acquisition operation that acquires speed information of the vehicle through the communication circuit. Claim 18 In claim 14, the collision risk assessment operation calculates the braking distance using vehicle speed information, road surface condition information, and slope information, and evaluates the collision risk using this value and the time to collision (TTC), an automatic emergency braking control method for a vehicle.

Images