Method and device for predicting vehicle path

Abstract

A vehicle path prediction method performed by a processor is disclosed. The vehicle path prediction method comprises the steps of: selecting surrounding vehicles from frames generated by a sensor; selecting the surrounding vehicle with the shortest collision time with the vehicle among the selected surrounding vehicles; modifying the path of the vehicle using the velocity vector of the selected surrounding vehicle; and re-modifying the path using three points and a Bézier curve. The three points include a first point representing the position of the vehicle, a second point representing the point where different line segments forming the modified vehicle path intersect, and a third point representing a position located at an arbitrary distance from the second point. The Bézier curve is expressed as shown in the following mathematical formula. The above B(t) represents the modified path, the above PA represents the first point, the above PB represents the second point, and the above PC represents the third point.

Description

Technology Field

[0001] The present invention relates to a vehicle path prediction method and apparatus, and more specifically, to a vehicle path prediction method and apparatus for preventing collisions of vehicles. Background Technology

[0002] Kinematic models are used to predict the movement path of a moving vehicle. A kinematic model is a mathematical algorithm that describes a vehicle's movement using its position, velocity, acceleration, and time. However, when a vehicle's left or right turn path is predicted using a kinematic model, the predicted path at a future point in time may be inaccurate. When the predicted path is inaccurate, it can lead to issues with safe driving, such as collisions with surrounding vehicles. Therefore, a new method is required to predict the vehicle's movement path more accurately. Prior art literature

[0003] Korean Published Patent Application No. 10-2024-0166171 (Nov. 26, 2024) The problem to be solved

[0004] The technical problem that the present invention aims to solve is to provide a vehicle path prediction method and device for preventing collisions with surrounding vehicles. means of solving the problem

[0005] A vehicle path prediction method performed by a processor according to an embodiment of the present invention includes the steps of selecting surrounding vehicles from frames generated by a sensor, selecting the surrounding vehicle with the shortest collision time with the vehicle among the selected surrounding vehicles, and modifying the path of the vehicle using the velocity vector of the selected surrounding vehicle.

[0006] The step of selecting surrounding vehicles in frames generated by the sensor comprises: detecting surrounding vehicles in the frames; predicting the kinematics circle of the vehicle; preliminarily selecting surrounding vehicles located outside the predicted kinematics circle of the vehicle from the detected surrounding vehicles; and finally selecting surrounding vehicles having the same direction as the vehicle's planned direction of travel from among the preliminarily selected surrounding vehicles.

[0007] A computing device according to an embodiment of the present invention includes a processor that executes vehicle path prediction commands and a memory that stores said commands.

[0008] The above commands are implemented to select surrounding vehicles from frames generated by the sensor, select the surrounding vehicle with the shortest collision time with the vehicle among the selected surrounding vehicles, and modify the vehicle's path using the velocity vector of the selected surrounding vehicle.

[0009] Commands for selecting surrounding vehicles in frames generated by the sensor are implemented to detect surrounding vehicles in the frames, predict the kinematics circle of the vehicle, preliminarily select surrounding vehicles located outside the predicted kinematics circle of the vehicle among the detected surrounding vehicles, and finally select surrounding vehicles having the same direction as the vehicle's planned direction of travel from among the preliminarily selected surrounding vehicles. Effects of the invention

[0010] The vehicle path prediction method and device according to an embodiment of the present invention have the effect of more accurately predicting collisions between a vehicle and surrounding vehicles by modifying the vehicle's path using the velocity vectors of surrounding vehicles. Brief explanation of the drawing

[0011] Detailed descriptions of each drawing are provided to help to more fully understand the drawings cited in the detailed description of the present invention. Figure 1 shows a block diagram of a vehicle path prediction system according to an embodiment of the present invention. Figure 2 shows a conceptual diagram to explain the problems that may occur when predicting a vehicle's path using a conventional kinematics model. Figure 3 shows a conceptual diagram for explaining a conventional kinematics model. FIGS. 4 to 6 show conceptual diagrams for explaining a vehicle path prediction method according to an embodiment of the present invention. FIG. 7 shows a conceptual diagram for explaining a vehicle path prediction method according to another embodiment of the present invention. FIG. 8 shows a flowchart for explaining the operation of a vehicle path prediction method according to an embodiment of the present invention. Specific details for implementing the invention

[0012] Specific structural or functional descriptions regarding embodiments according to the concept of the present invention disclosed herein are provided merely for the purpose of explaining embodiments according to the concept of the present invention, and embodiments according to the concept of the present invention may be implemented in various forms and are not limited to the embodiments described herein.

[0013] Embodiments according to the concept of the present invention may be subject to various modifications and may take various forms; therefore, embodiments are illustrated in the drawings and described in detail in this specification. However, this is not intended to limit the embodiments according to the concept of the present invention to specific disclosed forms, and includes all modifications, equivalents, or substitutions that fall within the spirit and scope of the present invention.

[0014] Terms such as "first" or "second" may be used to describe various components, but said components should not be limited by said terms. For the sole purpose of distinguishing one component from another, for example, without departing from the scope of rights according to the concept of the present invention, the first component may be named the second component, and similarly, the second component may be named the first component.

[0015] When it is stated that one component is "connected" or "connected" to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. Conversely, when it is stated that one component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between. Other expressions describing the relationship between components, such as "between" and "exactly between," or "adjacent to" and "directly adjacent to," should be interpreted in the same way.

[0016] The terms used herein are used merely to describe specific embodiments and are not intended to limit the invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this specification, terms such as “comprising” or “having” are intended to indicate the existence of the described features, numbers, steps, actions, components, parts, or combinations thereof, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0017] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the present invention pertains. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this specification.

[0018] Hereinafter, the present invention will be described in detail by explaining preferred embodiments of the present invention with reference to the attached drawings.

[0019] Figure 1 shows a block diagram of a vehicle path prediction system according to an embodiment of the present invention.

[0020] Referring to FIG. 1, the vehicle path prediction system (100) means a system for predicting the path of a vehicle (20).

[0021] The system (100) includes a lidar sensor (21) and a computing device (10).

[0022] A LiDAR sensor (21) is installed on a moving vehicle (20). The vehicle (20) may be an autonomous vehicle. The LiDAR sensor (21) emits laser light from the environment toward an object (e.g., vehicle (30)) to generate a 3D point cloud (25). The 3D point cloud (25) includes a set of LiDAR points. The set of LiDAR points may be referred to as LiDAR point data (25). The set of LiDAR points includes a plurality of LiDAR points. In FIG. 1, reference numeral 23 indicates a LiDAR point. A LiDAR point (23) may be expressed in (x, y, z) 3D coordinates. In this specification, LiDAR point, LiDAR point data, 3D point cloud, or LiDAR point cloud may be understood as the same concept.

[0023] Vehicles (20, 30) drive on the road (101). Vehicle (30) refers to surrounding vehicles.

[0024] The computing device (10) may be implemented as a hardware module combined with other hardware inside the vehicle (20), or as an independent hardware device. For example, the computing device (10) may be implemented in the Electronic Control Unit (ECU) of the vehicle (20). Additionally, the computing device (10) may be implemented as an external electronic device such as a computer, laptop, PC, server, or tablet PC.

[0025] A computing device (10) includes a processor (11) and a memory (13). The processor (11) executes instructions for predicting a vehicle path. The memory (13) stores the instructions. The operations of the computing device (10) can be understood as the operations of the processor (11).

[0026] The processor (11) receives a 3D point cloud (25) from the lidar sensor (21). Each lidar point included in the 3D point cloud (25) can be represented by 3D coordinates (x, y, z).

[0027] FIG. 2 shows a conceptual diagram to explain the problems that may occur when predicting the path of a vehicle using a conventional kinematics model. FIG. 2 (a) and FIG. 2 (b) show the actual path and predicted path of the vehicle (20).

[0028] In Fig. 2, the reference numeral '109' represents a road.

[0029] In FIG. 2, the reference numeral '103' indicates an object (e.g., a curbstone) other than the road (109).

[0030] When predicting the path of a moving vehicle (20) using a conventional kinematic model, the predicted path (105) of the vehicle (20) and the actual path (107) may differ. When the vehicle (20) turns left or right, the change in speed and angular velocity of the vehicle (20) is large. Therefore, in FIG. 2 (a) and FIG. 2 (b), the predicted path (105) of the vehicle (20) does not pass through the road (109) and may differ from the actual path (107).

[0031] Figure 3 shows a conceptual diagram for explaining a conventional kinematics model.

[0032] Referring to FIG. 3, a conventional kinematics model may be referred to as car kinematics. A conventional kinematics model means predicting the movement of a vehicle (20) using the position, speed, angular velocity, and time of the vehicle (20). The radius of rotation (R) is derived using the angular velocity (w) and speed (v) of the vehicle (20). Information regarding the position, speed, angular velocity, and time of the vehicle (20) may be generated by an IMU sensor or an INS (Inertial Navigation Systems) sensor included in the vehicle (20).

[0033] In FIG. 3, a path (105-1, 105-2) can be predicted using the radius of rotation (R). The path (105-1, or 105-2) is predicted based on the position, speed, angular velocity, and time of the vehicle (20). The path (105-1, 105-2) refers to the predicted path in FIG. 2.

[0034] The distance (d) is generated by multiplying the radius of rotation (R) and the angle (θ). The radius of rotation (R) is equal to the distance (d) divided by the angle (θ). This is given by the following mathematical formula.

[0035] [Mathematical Formula 1]

[0036]

[0037] The above t represents time. Therefore, the turning radius (R) can be calculated by dividing the speed (v) of the vehicle (20) by the angular velocity (w). This can be expressed as shown in the following mathematical formula.

[0038] [Mathematical Formula 2]

[0039]

[0040] FIGS. 4 to 6 show conceptual diagrams for explaining a vehicle path prediction method according to an embodiment of the present invention.

[0041] Referring to FIGS. 1 and FIGS. 4, the processor (11) selects surrounding vehicles (30, 40, 50) from frames generated by the lidar sensor (21). A frame refers to a 3D map of a scene generated from lidar points (23). A frame includes lidar points (23). Frames can be generated by the lidar sensor (21). The system (100) may have frame rates between 10 and 30 frames per second. According to an embodiment, a frame may be represented as a 2D map. Specific operations for selecting surrounding vehicles (30, 40, 50) may be performed as follows.

[0042] The processor (11) detects surrounding vehicles (30, 40, 50, 60, 70, 80, and 90) in the frames. A K-means algorithm or a deep learning algorithm may be used to detect surrounding vehicles (30, 40, 50, 60, 70, 80, and 90) in the frames.

[0043] The processor (11) predicts the kinematics circle (27) of the vehicle (20) using the position, speed, angular velocity, and time of the vehicle (20). The kinematics circle (27) refers to a path predicted using the conventional kinematics model described in FIG. 3. The vehicle (20) refers to a vehicle in motion. A LiDAR sensor (21) is installed on the vehicle (20) in motion.

[0044] The processor (11) preliminarily selects surrounding vehicles (30, 40, 50, 60, 70) located outside the predicted kinematic circle (27) of the vehicle (20) from the detected surrounding vehicles (30, 40, 50, 60, 70, 80, and 90). Outside the kinematic circle (27) means outside the circle when drawing a circle with respect to the center point (111). The kinematic circle (27) is an arc of the circle.

[0045] The processor (11) finally selects surrounding vehicles (30, 40, 50) that have the same direction as the planned direction of travel of the vehicle (20) from among the preliminarily selected surrounding vehicles (30, 40, 50, 60, 70). The processor (11) can predict or confirm the direction of the surrounding vehicles (30, 40, 50) in frames generated by the lidar sensor (21).

[0046] The processor (11) determines the intended direction of travel of the vehicle (20) (e.g., left turn, right turn, or straight direction) based on the position, speed, angular velocity, and time of the vehicle (20). When the intended direction of travel of the vehicle (20) is a right turn, the direction corresponding to the intended direction of travel of the vehicle (20) is the right direction. When the intended direction of travel of the vehicle (20) is a left turn, the direction corresponding to the intended direction of travel of the vehicle (20) is the left direction. When the intended direction of travel of the vehicle (20) is a straight direction, the direction corresponding to the intended direction of travel of the vehicle (20) is the upward direction.

[0047] When the planned direction of travel of the vehicle (20) is a right turn, the processor (11) finally selects surrounding vehicles (30, 40, 50) that have the same direction (right direction) as the planned direction of travel (right turn direction) of the vehicle (20).

[0048] After selecting surrounding vehicles (30, 40, 50), the processor (11) selects the surrounding vehicle (30) among the selected surrounding vehicles (30, 40, 50) that has the shortest collision time with the vehicle (20). The processor (11) can predict or verify the velocity vectors (31, 41, 51) of the surrounding vehicles (30, 40, 50) in frames generated by the lidar sensor (21).

[0049] In FIG. 4, reference numeral '31' represents the speed vector of a surrounding vehicle (30), reference numeral '41' represents the speed vector of a surrounding vehicle (40), and reference numeral '51' represents the speed vector of a surrounding vehicle (50). In FIG. 4, the arrows represent the speed vectors of the vehicles.

[0050] The processor (11) can calculate the collision time between the vehicle (20) and the selected surrounding vehicles (30, 40, 50) using the distance and velocity vectors (31, 41, 51) between the vehicle (20) and the selected surrounding vehicles (30, 40, 50). The collision time is expressed as shown in the following mathematical formula.

[0051] [Mathematical Formula 3]

[0052] TTC(Time to collision)=distance / velocity

[0053] The above TTC means the collision time, the above distance means the distance between the vehicle (20) and the surrounding vehicle (30, 40, or 50), and the above velocity means the speed of the surrounding vehicle (30, 40, or 50).

[0054] Since the number of selected surrounding vehicles (30, 40, 50) is 3, 3 collision times are calculated. The processor (11) selects the surrounding vehicle (30) that has the shortest collision time with vehicle (20) among the selected surrounding vehicles (30, 40, 50).

[0055] The processor (11) modifies the path (27) of the vehicle (20) using the direction of the velocity vector (31) of the selected surrounding vehicle (30). The modified path is represented by the reference numeral '29' in FIG. 5. Since the direction of the velocity vector (31) of the surrounding vehicle (30) is to the right, the processor (11) modifies the path (27) of the vehicle (20) from an arc to a straight line (29).

[0056] FIG. 7 shows a conceptual diagram for explaining a vehicle path prediction method according to another embodiment of the present invention.

[0057] Referring to FIG. 7, a method for more accurately predicting a modified path (29) is disclosed. A processor (11) has three points (P A , P B , and P C The path (37) is predicted more accurately using ). The processor (11) uses three points (P A , P B , and P C The path (29) is re-modified using ).

[0058] Point (P A ) represents the location of the moving vehicle (20). Point (P A ) may mean the center position of the moving vehicle (20). Point (P B ) refers to the point where different line segments (33, 35) forming the modified path (29) meet. Point (P B ) may mean the intersection point of the velocity vectors of the vehicle (20). point (P C ) is point(P B It means a location located at an arbitrary distance from ).

[0059] To predict the path (37) more accurately, a formula for a Bezier curve may be used. The formula for a Bezier curve is expressed as shown in the following mathematical equation.

[0060] [Mathematical Formula 4]

[0061]

[0062] The above B(t) is the re-modified path (37), and the above P A is the location of the moving vehicle (20), the above P B The point where different line segments (33, 35) meet in the modified path (29) is the P C is point (P B It refers to a position located at an arbitrary distance from ). In this case, t is a parameter representing the position according to B(t), and has a value between 0 and 1.

[0063] The processor (11) has three points (P A , P B , and P C The path (29) can be re-corrected using the formula for the Bezier curve. The processor (11) can predict the path of the vehicle (20) more accurately by generating the re-corrected path (37).

[0064] FIG. 8 shows a flowchart for explaining the operation of a vehicle path prediction method according to an embodiment of the present invention.

[0065] Referring to FIGS. 1 through 8, the processor (11) selects surrounding vehicles (30, 40, 50) from frames generated by the lidar sensor (21) (S10). The lidar sensor (21) is installed on a moving vehicle (20).

[0066] The processor (11) detects surrounding vehicles (30, 40, 50, 60, 70, 80, and 90) in the frames (S11).

[0067] The processor (11) predicts the kinematics circle of the vehicle (20) (S13).

[0068] The processor (11) preliminarily selects surrounding vehicles (30, 40, 50, 60, 70) located outside the predicted kinematic circle (27) of the vehicle (20) from the detected surrounding vehicles (30, 40, 50, 60, 70, 80, and 90) (S15).

[0069] The processor (11) finally selects surrounding vehicles (30, 40, 50) that have the same direction as the planned direction of travel of the vehicle (20) from among the preliminarily selected surrounding vehicles (30, 40, 50, 60, 70) (S17).

[0070] The processor (11) selects the surrounding vehicle (30) that has the shortest collision time with the vehicle (20) among the selected surrounding vehicles (30, 40, 50) (S20).

[0071] The processor (11) modifies the path (29) of the vehicle (20) using the speed vector of the selected surrounding vehicle (30) (S30).

[0072] The processor (11) is the location (P) of the vehicle (20). A Selected points (P) in the ) and modified path (29) B , P C Using ), the above modified path (29) is modified again to create a modified path (37) (S40).

[0073] The present invention has been described with reference to an exemplary embodiment illustrated in the drawings, but this is merely illustrative, and those skilled in the art will understand that various modifications and equivalent alternative embodiments are possible therefrom. Accordingly, the true technical scope of protection of the present invention should be determined by the technical spirit of the appended claims. Explanation of the symbols

[0074] 100: System; 10: Computing device; 11: Processor; 13: Memory; 20, 30: Vehicle; 21: LiDAR sensor; 25: LiDAR point data;

Claims

Claim 1 A method for predicting a vehicle path performed by a processor comprises: selecting surrounding vehicles from frames generated by a sensor; predicting a path including an arc using the turning radius of a vehicle traveling in a left-turn or right-turn direction; selecting the surrounding vehicle among the selected surrounding vehicles that has the shortest collision time with the vehicle; modifying the path of the vehicle using the velocity vector of the selected surrounding vehicle, wherein at least a portion of the arc is modified into a line corresponding to the planned direction of travel of the vehicle; and re-modifying the modified path using three points and a Bézier curve, wherein the three points include a first point representing the position of the vehicle, a second point representing the point where the line corresponding to the arc and the line corresponding to the planned direction of travel of the vehicle intersect in the modified path of the vehicle, and a third point representing a position located at an arbitrary distance from the second point, and wherein the Bézier curve is expressed as shown in the following mathematical formula. The above B(t) is the above modified path, and the above P A is the above-mentioned first point, the above P B is the above second point, the above P C is a vehicle path prediction method meaning the above-mentioned third point. Claim 2 A processor for executing vehicle path prediction commands; and a memory for storing said commands, wherein the commands are implemented to select surrounding vehicles from frames generated by a sensor, predict a path including an arc using the turning radius of a vehicle traveling in a left-turn or right-turn direction, select the surrounding vehicle among the selected surrounding vehicles that has the shortest collision time with the driving vehicle, modify the path of the driving vehicle using the velocity vector of the selected surrounding vehicle, modify at least a portion of the arc to a line corresponding to the driving vehicle's planned direction of travel, and re-modify the modified path using three points and a Bézier curve, wherein the three points include a first point representing the position of the driving vehicle, a second point representing the point where the line corresponding to the arc in the modified vehicle path meets the line corresponding to the driving vehicle's planned direction of travel, and a third point representing a position located at an arbitrary distance from the second point, and wherein the Bézier curve is expressed as shown in the following mathematical formula. The above B(t) is the above modified path, and the above P A is the above-mentioned first point, the above P B is the above second point, the above P C is a computing device meaning the above-mentioned third point.

Images