Method and device for predicting vehicle path based on intention of driver

Abstract

A method and apparatus for path prediction in a vehicle may be used to determine a driving intent of a driver. The vehicle path prediction method comprises the following steps executed by a processor: acquiring driving information of a vehicle; determining a longitudinal driving intention and a transverse driving intention of the driver based on the driving information; generating a speed distribution and a curvature distribution of the vehicle based on the longitudinal driving intention and the transverse driving intention; and determining a predicted path of the vehicle based on the velocity distribution and the curvature distribution.

Description

Methods and apparatus for vehicle route prediction based on driver intent Technical Field

[0001] This disclosure relates to a method and apparatus for performing vehicle path prediction based on driver intent, and more specifically, to a method and apparatus for improving path prediction accuracy when the vehicle cannot obtain lane information. Background Technology

[0002] Autonomous driving systems typically predict the vehicle's path for route planning, collision avoidance decisions, and other purposes. For example, an autonomous driving system can use the predicted path to calculate potential collisions between the autonomous vehicle and nearby objects, warn the driver, and execute evasive control measures.

[0003] If an autonomous vehicle can detect the lane of the road it is traveling on, the autonomous driving system can improve the accuracy of path prediction based on lane information. On the other hand, if the autonomous vehicle cannot detect the lane, existing autonomous driving systems will use the vehicle's driving information, including speed, acceleration, yaw rate, and steering angle, to predict the vehicle's path based on a kinematic model.

[0004] When lanes cannot be detected, and the prediction time window is long (e.g., exceeding 4 seconds), existing path prediction methods calculate the lateral movement and heading angle changes of the autonomous vehicle as it moves away from the current time point greater than the driver's driving intention. This means that existing path prediction methods have limitations in predicting paths over long prediction time windows when lanes cannot be detected. Summary of the Invention

[0005] This disclosure aims to provide a method and apparatus for performing path prediction with high accuracy even when the autonomous driving system is unable to perceive the lane.

[0006] This disclosure also aims to provide a method and apparatus for performing path prediction over a longer prediction time window, even when the autonomous driving system is unable to perceive the lane.

[0007] Furthermore, this disclosure provides a method and apparatus that can improve path prediction accuracy by performing path prediction for autonomous vehicles based on the driver's driving intentions.

[0008] The technical objectives to be achieved by this disclosure are not limited to those described above, and those skilled in the art can clearly understand other technical objectives not mentioned above through the detailed description below.

[0009] According to this disclosure, a method for predicting a vehicle path includes: acquiring vehicle driving information via at least one processor of the vehicle, wherein the driving information includes at least one of the vehicle's speed, acceleration, steering angle, steering angular velocity, heading angle, yaw rate, accelerator pedal and brake pedal depressure amount, or gear setting; determining a driver's longitudinal driving intention via at least one processor based on the driving information; determining a driver's lateral driving intention via at least one processor based on the driving information; generating a vehicle speed distribution via at least one processor based on the longitudinal and lateral driving intentions; generating a vehicle curvature distribution via at least one processor based on the longitudinal and lateral driving intentions; and determining a predicted path for the vehicle via at least one processor based on the speed distribution and curvature distribution.

[0010] According to another aspect of this disclosure, a method for predicting a vehicle path includes: acquiring vehicle driving information, wherein the driving information includes at least one of the vehicle's speed, acceleration, steering angle, steering angular velocity, heading angle, yaw rate, accelerator pedal and brake pedal depressure amount, or gear setting; determining a driver's longitudinal driving intention based on the driving information; determining a driver's lateral driving intention based on the driving information; generating a vehicle speed distribution based on the longitudinal and lateral driving intentions; generating a vehicle curvature distribution based on the longitudinal and lateral driving intentions; and determining a predicted path for the vehicle based on the speed distribution and curvature distribution.

[0011] According to another embodiment, this disclosure provides an apparatus for predicting a vehicle path, comprising: at least one memory configured to store instructions; and at least one processor, wherein the at least one processor executes instructions to cause the at least one processor to: acquire vehicle driving information, wherein the driving information includes at least one of the vehicle's speed, acceleration, steering angle, steering angular velocity, heading angle, yaw rate, accelerator pedal and brake pedal depressure amount, or gear setting; determine a driver's longitudinal driving intention based on the driving information; determine a driver's lateral driving intention based on the driving information; generate a vehicle speed distribution based on the longitudinal driving intention and the lateral driving intention; generate a vehicle curvature distribution based on the longitudinal driving intention and the lateral driving intention; and determine a predicted path for the vehicle based on the speed distribution and the curvature distribution.

[0012] According to at least one embodiment of this disclosure, a driving curve is generated by identifying the driver's longitudinal and lateral driving intentions, enabling the autonomous vehicle to perform path prediction with high accuracy even when the autonomous vehicle cannot perceive the lane.

[0013] According to at least one embodiment of the present disclosure, a driving curve is generated by identifying the driver's longitudinal and lateral driving intentions, enabling high-precision path prediction even for long prediction time windows.

[0014] Vehicles (e.g., autonomous vehicles) may include the aforementioned devices.

[0015] According to another aspect of this disclosure, a non-transitory computer-readable medium comprising program instructions executable by a processor includes: program instructions for acquiring vehicle driving information, wherein the driving information includes at least one of the following: vehicle speed, acceleration, steering angle, steering angular velocity, heading angle, yaw rate, accelerator pedal and brake pedal depressure amount, or gear setting; program instructions for determining a driver's longitudinal driving intention based on the driving information; program instructions for determining a driver's lateral driving intention based on the driving information; program instructions for generating a vehicle speed distribution based on the longitudinal and lateral driving intentions; program instructions for generating a vehicle curvature distribution based on the longitudinal and lateral driving intentions; and program instructions for determining a predicted path for the vehicle based on the speed distribution and curvature distribution.

[0016] In some respects, the vehicles described herein may be autonomous vehicles.

[0017] In fully autonomous vehicles or systems, the vehicle can perform all driving tasks under all conditions with little or no need for human driver assistance. For example, in semi-autonomous vehicles, the autonomous driving system can perform some or all driving tasks under certain conditions, but the human driver regains control under other conditions; or in other semi-autonomous systems, the vehicle's autonomous driving system can supervise steering, acceleration, and braking under certain conditions, but the human driver needs to continuously monitor the driving environment throughout the journey while performing the remaining necessary tasks.

[0018] In some embodiments, the system and vehicle may be fully autonomous. In other embodiments, the system and vehicle may be semi-autonomous.

[0019] The beneficial effects of this disclosure are not limited to those described above; those skilled in the art will clearly understand from the following description other beneficial effects of this disclosure not mentioned above. Attached Figure Description

[0020] Figure 1 is a schematic block diagram of a path prediction apparatus according to at least one embodiment of the present disclosure.

[0021] Figure 2 shows an example state transition diagram of a path prediction device according to at least one embodiment of the present disclosure for determining lateral driving intention.

[0022] Figures 3A to 3C are diagrams illustrating how the distribution generation unit determines the speed distribution of the vehicles.

[0023] Figures 4A to 4C illustrate how the distribution generation unit determines the curvature distribution of the vehicle.

[0024] Figure 5 is a flowchart illustrating the process by which a path prediction apparatus according to at least one embodiment of the present disclosure generates a predicted path for a vehicle.

[0025] Figures 6A and 6B are diagrams used to compare predicted paths generated by a path prediction method according to at least one embodiment of the present disclosure with predicted paths generated by conventional path prediction methods.

[0026] Figure 7 is a schematic block diagram of an exemplary computing device that can be used to implement the method or apparatus according to the present disclosure. Detailed Implementation

[0027] It should be understood that the terms "vehicle" or "of a vehicle" or other similar terms used herein encompass general motor vehicles, such as passenger cars (including SUVs), buses, trucks, various commercial vehicles, ships (including various vessels), aircraft, etc., as well as hybrid vehicles, electric vehicles, plug-in hybrid electric vehicles, hydrogen-powered vehicles, and other alternative fuel vehicles (e.g., fuels derived from resources other than petroleum). Hybrid vehicles, as referred to herein, are vehicles with two or more power sources, such as gasoline-powered and electric-powered vehicles.

[0028] The terminology used herein is for describing particular embodiments only and is not intended to limit the scope of this disclosure. The singular forms “a,” “an,” and “the” used herein also include the plural forms unless the context clearly indicates otherwise. Further understanding is that the terms “comprising” and / or “including” as used herein mean 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 combinations thereof. The term “and / or” as used herein includes any and all combinations of one or more associated listed items. Throughout this specification, unless otherwise expressly stated, the word “comprising” and variations thereof (e.g., “including” or “covering”) should be understood to imply the inclusion of the stated elements, but do not exclude any other elements. Furthermore, the terms “unit,” “device,” “piece,” and “module” described in the specification refer to a unit for performing at least one function and operation, and may be implemented by hardware components or software components and combinations thereof.

[0029] Furthermore, the control logic of this disclosure can be implemented as a non-transitory computer-readable medium containing executable program instructions that are executed by a processor, controller, etc. Examples of computer-readable media include, but are not limited to, ROM, RAM, optical disc (CD-ROM), magnetic tape, floppy disk, flash drive, smart card, and optical data storage device. The computer-readable medium can also be distributed across a network-connected computer system for distributed storage and execution, for example, via a remote information processing server or a controller area network (CAN).

[0030] Some exemplary embodiments of this disclosure will now be described in detail with reference to the accompanying drawings. In the following description, the same reference numerals preferably denote the same elements, even though these elements are shown in different drawings. Furthermore, for clarity and brevity, detailed descriptions of known functions and configurations incorporated therein will be omitted in the description of some of the following embodiments.

[0031] Furthermore, terms such as first, second, A, B, (a), (b) are used only to distinguish one component from another and do not imply or suggest the composition, order, or sequence of components. In this specification, when a component “comprises” or “includes” another component, that component is intended to further include other components, not exclude other components, unless otherwise expressly stated. Terms such as “unit” or “module” refer to one or more units for performing at least one function or operation, which may be implemented by hardware, software, or a combination thereof.

[0032] The following detailed description and accompanying drawings are intended to describe exemplary embodiments of the present disclosure and are not intended to represent the only embodiments that may be practiced with respect to the present disclosure.

[0033] As used herein, path prediction refers to the function of an autonomous driving system in predicting the path of a vehicle, which may be an autonomous vehicle (also referred to as "the vehicle" in this disclosure).

[0034] In this disclosure, the prediction time window refers to the time interval from the current time point to the future time point that the autonomous driving system wants to predict. The length of the prediction time window can be expressed in units of time. For example, if the autonomous driving system predicts the path from the current time point to a future time point 4 seconds later, then the length of the prediction time window is 4 seconds.

[0035] Figure 1 is a schematic block diagram of a path prediction apparatus 100 according to at least one embodiment of the present disclosure.

[0036] The path prediction device 100 includes a memory 110 and a processor 120. The path prediction device 100 can be implemented as an embedded device, a server, electronics within an autonomous driving system, etc. Not all modules shown in Figure 1 are essential components; some modules included in the path prediction device 100 can be added, modified, or removed. Furthermore, the components shown in Figure 1 represent elements categorized by function, and at least one component can be implemented in an integrated form in a real physical environment.

[0037] The memory 110 stores the data and commands required for the operation of the path prediction device 100.

[0038] The memory 110 can store vehicle driving information and road information obtained by using at least one sensor included in the vehicle. The vehicle driving information may include vehicle speed, acceleration, steering angle, steering angular rate, heading angle, yaw rate, accelerator / brake pedal engagement and / or gear setting.

[0039] Processor 120 controls the overall operation of path prediction device 100. Processor 120 can be implemented as one or more processors. Processor 120 can execute instructions stored in memory 110.

[0040] The processor 120 may include a driving intention determination unit 122, a distribution generation unit 124, and a path generation unit 126.

[0041] The driving intention determination unit 122 can determine the driver's driving intention based on the vehicle's driving information stored in the memory 110. The driver's driving intention can be divided into longitudinal driving intention and lateral driving intention. Longitudinal driving intention refers to the driver's intention to drive the vehicle at high or low speed, to accelerate or decelerate rapidly, etc. Lateral driving intention refers to the driver's intention to turn the vehicle at a certain angle, to drive straight or not straight, to make a U-turn or not, etc.

[0042] In at least one embodiment of this disclosure, longitudinal driving intentions can be categorized as idling, low speed, high speed, rapid acceleration / deceleration, etc. Table 1 describes the definitions of different types of longitudinal driving intentions.

[0043] [Table 1]

[0044]

[0045] If the vehicle is in D or R gear and there is no input from the accelerator or brake pedal, the driving intention determination unit 122 determines the driver's longitudinal driving intention to be Idle. If the vehicle's acceleration is less than or equal to a predetermined acceleration a... refSize, and vehicle speed less than or equal to the predetermined speed v ref Then, the driving intention determination unit 122 determines the driver's longitudinal driving intention as LowSpd. If the vehicle's acceleration is less than or equal to the predetermined acceleration a... ref Size, and vehicle speed greater than predetermined speed v ref Then, the driving intention determination unit 122 determines the driver's longitudinal driving intention as HighSpd. If the vehicle's acceleration is greater than the predetermined acceleration a... ref If the size is determined, the driving intention determination unit 122 determines the driver's longitudinal driving intention as RapidAcc.

[0046] In at least one embodiment of this disclosure, lateral driving intentions can be categorized as Restore, SmoothTurn, Unknown, Turn, Counter, etc. Table 2 describes the definitions of each type of lateral driving intention.

[0047] [Table 2]

[0048]

[0049] SmoothTurn represents the driver's intention to change the route to a hypothetical path parallel to the current driving route. This occurs if the vehicle's steering angle is less than or equal to the first predetermined steering angle θ. ref1 And / or if the vehicle's steering angular velocity is less than or equal to the first predetermined steering angular velocity The driving intention determination unit 122 then determines that the driver's lateral driving intention is SmoothTurn. The first predetermined steering angle and the first predetermined steering angular velocity are set based on the steering angle and steering angular velocity that occur during a smooth change in the route (e.g., lane change).

[0050] Restore signifies the driver's intention to complete the route change to the hypothetical path and align the vehicle with it. This occurs if the vehicle's steering angle is less than or equal to the first predetermined steering angle θ. ref1 Furthermore, the yaw rate of the vehicle is out of phase with the steering angle of the vehicle. That is, due to the phase lag of the yaw rate output, the yaw rate and the steering angle have opposite signs. Therefore, the driving intention determination unit 122 determines that the driver's lateral driving intention is Restore.

[0051] "Turn" indicates the driver's intention to steer towards an imaginary path that is not parallel to the driver's current path. If the vehicle's steering angle is greater than a second predetermined steering angle θ... ref2 And the vehicle's steering angular velocity is greater than the second predetermined steering angular velocity. The driving intention determination unit 122 then determines that the driver's lateral driving intention is Turn. The second predetermined steering angle and the second predetermined steering angular velocity are set based on the steering angle and steering angular velocity that occur during a sudden change in the route (e.g., a U-turn).

[0052] Counter indicates the driver's intention to correct oversteer input. If the vehicle's steering angle is greater than the second predetermined steering angle θ... ref2 Furthermore, the vehicle's steering angle and steering angular velocity are out of phase. That is, due to the phase lag of the steering angle output, the steering angle and steering angular velocity have opposite signs. Therefore, the driving intention determination unit 122 determines the driver's lateral driving intention as Counter.

[0053] "Unknown" indicates that it is unclear whether the driver's lateral intention is a smooth turn or a sharp turn. If the vehicle's steering angle is greater than the first predetermined steering angle θ... ref1 But less than or equal to the second predetermined steering angle θ ref2 And / or if the vehicle's steering angular velocity is greater than a first predetermined steering angular velocity But less than or equal to the second predetermined steering angular velocity Then, the driving intention determination unit 122 can determine that the driver's lateral driving intention is Unknown. In cases where it is difficult to determine whether the lateral driving intention is a Turn or a Smooth Turn, the driving intention determination unit 122 can determine the lateral driving intention as Unknown.

[0054] Since the driving intention determination unit 122 determines the driver's lateral driving intention based on a hypothetical path when lane detection is unavailable, a state transition diagram can be used to prevent unintentional operation by the driver.

[0055] For example, Restore requires the vehicle to have a small steering angle. Therefore, the state can transition from SmoothTurn to Restore, but not from Unknown, Turn, or Counter. As another example, Counter requires the vehicle to have a large steering angle. Therefore, the state can transition from Turn to Counter, but not from Unknown, SmoothTurn, or Restore.

[0056] Figure 2 illustrates an example state transition diagram for a path prediction apparatus 100 according to at least one embodiment of the present disclosure to determine lateral driving intention. At the start of lateral driving intention determination, the driver's lateral driving intention is Unknown.

[0057] If the vehicle's steering angle is less than or equal to the first predetermined steering angle θ ref1 And / or if the vehicle's steering angular velocity is less than or equal to the first predetermined steering angular velocity Then the driving intention determination unit 122 changes the lateral driving intention from Unknown to SmoothTurn. Conversely, if the vehicle's steering angle is greater than the first predetermined steering angle θ... ref1 And / or if the vehicle's steering angular velocity is greater than a first predetermined steering angular velocity Then the driving intention determination unit 122 will change the lateral driving intention from SmoothTurn to Unknown or Turn.

[0058] If the vehicle's steering angle is greater than the second predetermined steering angle θ ref2 And / or if the vehicle's steering angular velocity is greater than the second predetermined steering angular velocity Then the driving intention determination unit 122 changes the lateral driving intention from Unknown to Turn. Conversely, if the vehicle's steering angle is less than or equal to the second predetermined steering angle θ... ref2 And / or the vehicle's steering angular velocity is less than or equal to the second predetermined steering angular velocity. Then the driving intention determination unit 122 will change the lateral driving intention from Turn to Unknown or SmoothTurn.

[0059] If the yaw rate and steering angle have opposite signs when the lateral travel intention is SmoothTurn, the travel intention determination unit 122 can change the lateral travel intention from SmoothTurn to Restore. Conversely, if the yaw rate and steering angle have the same sign when the lateral travel intention is Restore, the travel intention determination unit 122 can change the lateral travel intention from Restore to SmoothTurn.

[0060] If, when the lateral driving intention is Turn, the steering angular velocity has the opposite sign to the steering angle, the driving intention determination unit 122 can change the lateral driving intention from Turn to Counter. Conversely, if, when the lateral driving intention is Counter, the steering angular velocity has the same sign as the steering angle, the driving intention determination unit 122 can change the lateral driving intention from Counter to Turn. Alternatively, if, when the lateral driving intention is Counter, the steering angle becomes less than or equal to the second predetermined steering angle θ... ref2 Then, the driving intention determination unit 122 can change the lateral driving intention from Counter to SmoothTurn or Unknown.

[0061] The distribution generation unit 124 can generate a speed distribution and a curvature distribution for predicting a vehicle path based on the driver's longitudinal driving intention and / or lateral driving intention.

[0062] In the present disclosure, the speed distribution is information representing how the vehicle speed changes over time, and the curvature distribution is information representing how the curvature of the vehicle trajectory changes over time. The speed distribution and the curvature distribution can be represented in the form of an array. For example, if the prediction time window is 4 seconds and the driving speed or the driving curvature is predicted at an interval of 0.1 second, the speed distribution or the curvature distribution can be composed of 41 elements.

[0063] In at least one embodiment of the present disclosure, the distribution generation unit 124 can generate a speed distribution based on the driver's longitudinal driving intention. In at least one embodiment of the present disclosure, the distribution generation unit 124 can adjust the speed distribution based on the driver's lateral driving intention. Various methods (such as multiplying by a correction factor, linear combination, etc.) can be used to reflect the lateral driving intention in the speed distribution.

[0064] FIGS. 3A to 3C are diagrams showing a method by which the distribution generation unit 124 determines the speed distribution of a vehicle through linear combination. In FIGS. 3A to 3C, the horizontal axis of the diagram represents future time points, the vertical axis of the diagram represents the driving speed of the vehicle, and T on the horizontal axis represents the length of the prediction time window.

[0065] FIG. 3A shows an example speed distribution diagram generated by the distribution generation unit 124 in a situation where the vehicle travels at a high speed and then decelerates.

[0066] In FIG. 3A, the first line represents the speed distribution with a longitudinal driving intention of HighSpd, and the second line represents the speed distribution with a longitudinal driving intention of RapidAcc. In a situation where the vehicle travels at a high speed and then decelerates, if the magnitude of the deceleration is less than or equal to a predetermined acceleration a ref , the driver's longitudinal driving intention is determined to be HighSpd, and if the magnitude of the deceleration is greater than the predetermined acceleration a ref , the driver's longitudinal driving intention is determined to be RapidAcc.

[0067] In FIG. 3A, the first deceleration interval (0 < t < t1) is an interval reflecting the deceleration when the longitudinal driving intention is HighSpd. The second deceleration interval (0 < t < t2) is an interval reflecting the deceleration when the longitudinal driving intention is RapidAcc. The distribution generation unit 124 can generate a speed distribution by setting the first deceleration interval and the second deceleration interval differently. For example, if the longitudinal driving intention is RapidAcc, the second deceleration interval can be set longer than the first deceleration interval to reflect the driver's intention to significantly reduce the vehicle speed.

[0068] In addition, based on the driver's lateral driving intention, the distribution generation unit 124 can adjust the initial speed distribution generated based on the driver's longitudinal driving intention. For example, if it is determined that the driver's lateral driving intention is Turn or Counter, the distribution generation unit 124 can generate the speed distribution in a manner that further increases the deceleration interval (i.e., increases t1 and t2).

[0069] FIG. 3B shows another example speed distribution diagram generated by the distribution generation unit 124 in a situation where the vehicle travels at a low speed and then decelerates.

[0070] In FIG. 3B, the first line represents the speed distribution with a longitudinal driving intention of LowSpd, and the second line represents the speed distribution with a longitudinal driving intention of RapidAcc. In a situation where the vehicle travels at a low speed and then decelerates, if the magnitude of the acceleration is less than or equal to a predetermined acceleration a ref , it is determined that the driver's longitudinal driving intention is LowSpd, and if the magnitude of the acceleration is greater than the predetermined acceleration a ref , it is determined that the driver's longitudinal driving intention is RapidAcc.

[0071] In FIG. 3B, the first deceleration interval (0 < t < t1) is the interval reflecting the deceleration when the longitudinal driving intention is LowSpd, and the second deceleration interval (0 < t < t2) is the interval reflecting the deceleration when the longitudinal driving intention is RapidAcc. The distribution generation unit 124 can generate the speed distribution by setting the first deceleration interval and the second deceleration interval differently. For example, if the longitudinal driving intention is RapidAcc, the second deceleration interval can be set longer than the first deceleration interval to reflect the driver's intention to significantly reduce the vehicle speed.

[0072] In addition, based on the driver's lateral driving intention, the distribution generation unit 124 can adjust the initial speed distribution generated based on the driver's longitudinal driving intention. For example, if it is determined that the driver's lateral driving intention is not Unknown, the distribution generation unit 124 can generate the speed distribution by increasing the deceleration interval (i.e., increasing t1 and t2).

[0073] FIG. 3C shows yet another example speed distribution diagram generated by the distribution generation unit 124 in a situation where the driver places the vehicle in the D gear or R gear and does not operate the accelerator pedal or brake pedal of the vehicle. In this situation, it is determined that the driver's lateral driving intention is Idle. If the lateral driving intention is Idle, the distribution generation unit 124 generates the speed distribution by depicting the speed characteristic curve during the creep driving. In FIG. 3C, the first line represents the speed distribution generated, for example, by depicting the speed characteristic curve during the creep driving.

[0074] In at least one embodiment of this disclosure, the distribution generation unit 124 can generate a curvature distribution based on the driver's lateral driving intention. In at least one embodiment of this disclosure, the distribution generation unit 124 can adjust the curvature distribution based on the driver's longitudinal driving intention. Various methods (e.g., multiplying by a correction factor, linear combination, etc.) can be used to reflect the longitudinal driving intention in the curvature distribution.

[0075] Figures 4A to 4C illustrate the method by which the distribution generation unit 124 determines the curvature distribution of the vehicle by multiplying by a correction coefficient. In Figures 4A to 4C, the horizontal axis represents a future time point, the vertical axis represents the vehicle's driving curvature, and T on the horizontal axis represents the length of the prediction time window.

[0076] Figure 4A shows an example curvature distribution generated by the distribution generation unit 124 when the driver's lateral driving intention is Restore. In Figure 4A, the first line represents the curvature distribution generated based on the lateral driving intention, and the second line represents the curvature distribution adjusted to reflect the longitudinal driving intention.

[0077] Since Restore represents the driver's intention to change the route to the hypothetical path and align the vehicle with the hypothetical path, the distribution generation unit 124 can reflect the driver's intention by generating a curvature distribution (the first line in FIG4A) with a curved decay (e.g., in the form of a quadratic function).

[0078] Furthermore, based on the driver's longitudinal driving intention, the distribution generation unit 124 can adjust the curvature distribution generated based on the driver's lateral driving intention. In at least one embodiment of this disclosure, the distribution generation unit 124 can multiply the curvature distribution generated based on the driver's lateral driving intention by a correction coefficient, thereby using the adjusted curvature distribution (the second line in FIG. 4A) as the curvature distribution. The value of the correction coefficient can be between 0 and 1. In one example, if the driver's longitudinal driving intention is LowSpd or Idle, the driving curvature may be large, so the correction coefficient can be set to a value close to 1. In another example, if the driver's longitudinal driving intention is HighSpd or RapidAcc, the correction coefficient can be set to a value close to 0, because the driving curvature is unlikely to be large. The correction coefficient based on the longitudinal driving intention can be determined empirically.

[0079] Figure 4B shows another example curvature distribution map generated by the distribution generation unit 124 when the driver's lateral driving intention is SmoothTurn. In Figure 4B, the first line represents the curvature distribution generated based on the lateral driving intention, and the second line represents the curvature distribution adjusted to reflect the longitudinal driving intention.

[0080] Since SmoothTurn is the intention to change the route to a hypothetical path parallel to the driver's current path, the distribution generation unit 124 can reflect the driver's intention by generating a curvature distribution (the first line in FIG4B) with linear decay (e.g., in the form of a linear function).

[0081] Furthermore, based on the driver's longitudinal driving intention, the distribution generation unit 124 can adjust the curvature distribution generated based on the driver's lateral driving intention. In at least one embodiment of this disclosure, the distribution generation unit 124 can multiply the curvature distribution generated based on the driver's lateral driving intention by a correction coefficient, thereby using the adjusted curvature distribution (the second line in FIG. 4B) as the curvature distribution. This correction coefficient can be the same as the correction coefficient used in FIG. 4A, or it can be a different value. The correction coefficient based on the longitudinal driving intention can be determined empirically.

[0082] Figure 4C shows another example curvature distribution map generated by the distribution generation unit 124 when the driver's lateral driving intention is Counter. In Figure 4C, the first line represents the curvature distribution generated based on the lateral driving intention, and the second line represents the curvature distribution adjusted to reflect the longitudinal driving intention.

[0083] Since the Counter is the driver’s intention to counteract the oversteering input, the distribution generation unit 124 can reflect the driver’s intention by generating a curvature distribution with the following shape (the first line in Figure 4C): constant curvature that then drops to 0 (e.g., in the form of a step function).

[0084] Furthermore, based on the driver's longitudinal driving intention, the distribution generation unit 124 can adjust the curvature distribution generated based on the driver's lateral driving intention. In at least one embodiment of this disclosure, the distribution generation unit 124 can multiply the curvature distribution generated based on the driver's lateral driving intention by a correction coefficient, thereby using the adjusted curvature distribution (the second line in FIG. 4C) as the curvature distribution. This correction coefficient can be the same as the correction coefficient used in FIG. 4A or FIG. 4B, or it can be a different value. The correction coefficient based on the longitudinal driving intention can be determined empirically.

[0085] If the driver's lateral driving intention is Turn or Unknown, the distribution generation unit 124 generates a curvature distribution with invariant curvature, that is, a curvature distribution with a constant form.

[0086] The path generation unit 126 predicts the vehicle path based on the velocity distribution and curvature distribution. In this disclosure, the predicted path is information indicating the changes in the vehicle's position (x, y) and heading angle (θ) over time, and can be represented in array form. For example, if the prediction time window is 4 seconds and the prediction time interval is 0.1 seconds, the predicted path can consist of 41 elements.

[0087] In at least one embodiment of this disclosure, the path generation unit 126 calculates the vehicle position and heading angle at each time point by using the vehicle position (X0, Y0) and heading angle (θ0) at the current time point, as well as the velocity distribution and curvature distribution, thereby predicting the vehicle's path.

[0088] Formula 1 is a path generation unit 126 using the vehicle position (Xt, Yt) and heading angle (θ) at time t. t To calculate the vehicle position (X) at time (t+1). t+1 ,Y t+1 ) and heading angle (θ) t+1 ).

[0089] [Formula 1]

[0090] X t+1 =X t +V t ·Δt·Cos(θ t )

[0091] Y t+1 =Y t +V t ·Δt·Sin(θ t )

[0092] θ t+1 =θ t +ρ t ·V t ·Δt

[0093] In Formula 1, X and Y represent the vehicle's position, θ represents the vehicle's heading angle, V represents the vehicle's velocity distribution, and ρ represents the vehicle's curvature distribution. Δt represents the prediction time interval. For example, if the prediction time window is 4 seconds and the prediction time interval is 0.1 seconds, then path generation unit 126 calculates (X0, Y0, θ0) to (X... 40 ,Y 40 ,θ 40 ).

[0094] Figure 5 is a flowchart illustrating the process by which a path prediction apparatus 100 generates a predicted path for a vehicle according to at least one embodiment of the present disclosure.

[0095] The path prediction device 100 uses at least one sensor included in the vehicle to obtain the vehicle's current driving information (operation S510). The vehicle's current driving information includes at least one of the following: vehicle speed, acceleration, steering angle, steering angular rate, heading angle, yaw rate, accelerator / brake pedal depressurization, idle speed information, or gear setting.

[0096] The path prediction device 100 determines the driver's longitudinal and lateral driving intentions based on the vehicle's current driving information (operation S520). In at least one embodiment, the path prediction device 100 can determine the longitudinal driving intention based on information such as the current vehicle speed, acceleration, pedal depressor, idle speed information, and gear setting. In at least one embodiment, the path prediction device 100 can determine the lateral driving intention based on information such as the vehicle's current steering angle, steering angular velocity, and yaw rate.

[0097] The path prediction device 100 generates a speed distribution and a curvature distribution based on the driver's longitudinal and lateral driving intentions (operation S530). In at least one embodiment, the path prediction device 100 can generate a speed distribution based on the driver's longitudinal driving intention and correct the speed distribution based on the lateral driving intention. In at least one embodiment, the path prediction device 100 can generate a curvature distribution based on the driver's lateral driving intention and correct the curvature distribution based on the longitudinal driving intention.

[0098] The path prediction device 100 predicts the vehicle path based on the velocity distribution and curvature distribution (operation S540). In at least one embodiment, the path prediction device 100 can use Formula 1 to predict the vehicle path.

[0099] Figures 6A and 6B are diagrams used to compare predicted paths generated by a path prediction method according to at least one embodiment of the present disclosure with predicted paths generated by conventional path prediction methods.

[0100] Figure 6A is a schematic diagram comparing the predicted path generated by the path prediction device 100 with the predicted path generated by a conventional path prediction method when the vehicle 10 turns left at an intersection. In Figure 6A, the first path represents the actual driving route of the vehicle, the second path represents the predicted path generated by the conventional path prediction device, and the third path represents the predicted path generated by the path prediction method of this disclosure.

[0101] Previous path prediction methods rely on kinematic models to predict paths when intersection conditions and lane information cannot be detected. In other words, path prediction is performed under the assumption that the vehicle 10's current speed, yaw rate, and other parameters remain unchanged at future points in time. Therefore, the generated path (second path) deviates significantly from the actual driving path (first path).

[0102] On the other hand, the path prediction device 100 of this disclosure generates a predicted path based on the driver's longitudinal and lateral driving intentions, thereby generating a path (third path) that does not deviate significantly from the actual driving path (first path). In Figure 6A, the driver's lateral driving intention is determined to be SmoothTurn, while the driver's longitudinal driving intention is determined to be LowSpd. In this case, as shown in Figure 4B, the curvature distribution exhibits a linear decay. Therefore, the predicted path (third path) generated by the path prediction method of this disclosure gradually approaches a straight line, thereby generating a path that does not deviate significantly from the actual driving path (first path).

[0103] Figure 6B is a schematic diagram comparing the predicted path generated by the path prediction device 100 with the predicted path generated by a conventional path prediction method when the vehicle 10 is traveling on a narrow road and avoiding obstacle 610. In Figure 6B, the first path represents the actual travel path of the vehicle, the second path represents the predicted path generated by the conventional path prediction device, and the third path represents the predicted path generated by the path prediction method of this disclosure. Figure 6B assumes that the vehicle 10 is traveling in a situation where the lane cannot be detected due to the obstacle.

[0104] Previous path prediction methods predicted paths based on kinematic models when lanes could not be detected. In other words, path prediction was performed under the assumption that the vehicle 10's current speed, yaw rate, and other parameters would remain constant at future points in time. Therefore, the generated path (the second path) deviated significantly from the actual driving path (the first path). If the autonomous vehicle made a collision determination based on the second path, it might determine that it had collided with another obstacle 620 located to its left front.

[0105] On the other hand, the path prediction device 100 of this disclosure generates a predicted path based on the driver's longitudinal and lateral driving intentions, thereby generating a path (third path) that does not deviate significantly from the actual driving path (first path). In Figure 6B, the driver's lateral driving intention is determined to be Restore, while the driver's longitudinal driving intention is determined to be LowSpd. In this case, as shown in Figure 3A, the curvature distribution exhibits an initial rapid decline. Therefore, the predicted path (third path) generated by the path prediction method of this disclosure only has a large curvature at the beginning and quickly approaches a straight line, thereby enabling the generation of a path that does not deviate significantly from the actual driving path (first path).

[0106] Figure 7 is a schematic block diagram of an illustrative configuration of a computing device 700 that can be used to implement the method or apparatus according to the present disclosure.

[0107] The computing device 700 may include some or all of the following: memory 710, processor 720, storage device 730, input / output interface 740, and communication interface 750. The computing device 700 may structurally and / or functionally include at least a portion of the path prediction device 100. The computing device 700 may be a fixed computing device, such as a desktop computer or server, or a mobile computing device, such as a laptop or smartphone. The computing device 700 may include any dedicated hardware accelerator capable of efficiently processing AI model calculations. For example, the computing device 700 may include a graphics processing unit (GPU), a tensor processing unit (TPU), or a neural processing unit (NPU).

[0108] Memory 710 may store programs that cause processor 720 to perform the methods or operations of the embodiments of this disclosure. For example, the program may include a plurality of instructions executable by processor 720, and processor 720 may execute these instructions to perform the methods or operations described above. Memory 710 may be a single memory or multiple memories. In this case, the information required to perform the methods or operations of the embodiments of this disclosure may be stored in a single memory or stored separately in multiple memories. When memory 710 consists of multiple memories, they may be physically separated. Memory 710 may include at least one of volatile memory or non-volatile memory. Volatile memory may include, for example, static random access memory (SRAM) or dynamic random access memory (DRAM), while non-volatile memory may include, for example, flash memory.

[0109] Processor 720 may include at least one core capable of executing at least one set of instructions. Processor 720 may execute instructions stored in memory 710. Processor 720 may be a single processor or multiple processors.

[0110] Even if the power supply to the computing device 700 is interrupted, the storage device 730 can retain the stored data. For example, the storage device 730 may include non-volatile memory, or storage media such as magnetic tape, optical disc, or magnetic disk. Programs stored in the storage device 730 can be loaded into the memory 710 before being executed by the processor 720. The storage device 730 can store files written in a programming language, and programs generated by a compiler or the like can be loaded into the memory 710 from these files. The storage device 730 can store data to be processed by the processor 720 and / or data already processed by the processor 720.

[0111] The input / output interface 740 can provide an interface with input devices (such as a keyboard, mouse, etc.) and / or output devices (such as a display device, printer, etc.). Users can trigger the processor 720 to execute programs via the input devices and / or view the processing results of the processor 720 via the output devices.

[0112] The communication interface 750 can provide access to external networks. The computing device 700 can communicate with other devices via the communication interface 750.

[0113] The various elements of the apparatus or method disclosed herein can be implemented in hardware, software, or a combination of hardware and software. The function of each element can be implemented in software, or the corresponding software function can be implemented by a microprocessor.

[0114] Various embodiments of the systems and techniques described herein can be implemented using digital electronic circuits, integrated circuits, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), computer hardware, firmware, software, and / or combinations thereof. Various embodiments may include one or more computer programs that execute on the programmable system. The programmable system includes at least one programmable processor, which may be a dedicated or general-purpose processor, connected to a storage system, at least one input device, and at least one output device, and receives and transmits data and instructions. The computer program (also referred to as a program, software, software application, or code) includes instructions for the programmable processor and is stored in a computer-readable recording medium.

[0115] Computer-readable recording media can include all types of storage devices capable of storing computer-readable data. Computer-readable recording media can be non-volatile or non-transitory media, such as read-only memory (ROM), random access memory (RAM), optical disc (CD-ROM), magnetic tape, floppy disk, or optical data storage devices. Furthermore, computer-readable recording media can also include transient media, such as data transmission media. Moreover, computer-readable recording media can be distributed across computer systems connected via a network, and computer-readable program code can be stored and executed in a distributed manner.

[0116] Although the operations shown in the flowcharts / timing diagrams in this specification are performed sequentially, this is merely an exemplary description of the technical concept of one embodiment of this disclosure. In other words, those skilled in the art to which this disclosure pertains will understand that various modifications and alterations can be made without departing from the essential characteristics of the embodiments of this disclosure; that is, the order shown in the flowcharts / timing diagrams can be changed, and one or more operations can be executed in parallel. Therefore, the flowcharts / timing diagrams are not limited to a chronological order.

[0117] While exemplary embodiments of this disclosure have been described for illustrative purposes, those skilled in the art will understand that various modifications, additions, and substitutions can be made without departing from the spirit and scope of the claims of this disclosure. Therefore, exemplary embodiments of this disclosure have been described for the sake of brevity and clarity. The scope of the technical concept of these embodiments is not limited by the description. Therefore, those skilled in the art will understand that the scope of the claims of this disclosure is not limited to the embodiments explicitly described above, but is limited by the claims and their equivalents.

Claims

1. A method for predicting vehicle paths, the method comprising the following steps: The system acquires vehicle driving information, including at least one of the following: vehicle speed, acceleration, steering angle, steering angular velocity, heading angle, yaw rate, accelerator pedal and brake pedal depressure, or gear setting; based on the driving information, it determines the driver's longitudinal driving intention; based on the driving information, it determines the driver's lateral driving intention; based on the longitudinal and lateral driving intentions, it generates a vehicle speed distribution; based on the longitudinal and lateral driving intentions, it generates a vehicle curvature distribution; and based on the speed distribution and the curvature distribution, it determines a predicted path for the vehicle.

2. The method according to claim 1, wherein, The step of determining the longitudinal driving intention includes: if the magnitude of the vehicle's acceleration is greater than a predetermined acceleration magnitude, then the longitudinal driving intention is determined to be a first longitudinal driving intention, wherein the first longitudinal driving intention is the driver's intention to accelerate or decelerate rapidly.

3. The method according to claim 2, wherein, The step of determining the longitudinal driving intention further includes: if the magnitude of the vehicle's acceleration is equal to or less than the predetermined acceleration magnitude, then determining whether the vehicle speed is equal to or less than the predetermined speed; and if the vehicle speed is equal to or less than the predetermined speed, then determining the longitudinal driving intention as a second longitudinal driving intention, and if the vehicle speed is greater than the predetermined speed, then determining the longitudinal driving intention as a third longitudinal driving intention, wherein the second longitudinal driving intention is the driver's intention to drive at a low speed, and the third longitudinal driving intention is the driver's intention to drive at a non-low speed.

4. The method according to claim 3, wherein, The step of determining the longitudinal driving intention further includes: if the vehicle is in D or R gear and there is no input from the accelerator pedal or brake pedal of the vehicle, then the longitudinal driving intention is determined to be a fourth longitudinal driving intention, wherein the fourth longitudinal driving intention is the driver's intention to crawl.

5. The method according to claim 4, wherein, The step of determining the lateral driving intention includes: if the vehicle's steering angle is equal to or less than a first predetermined steering angle, or if the vehicle's steering angular velocity is equal to or less than the first predetermined steering angular velocity, then the lateral driving intention is determined to be a first lateral driving intention; if the vehicle's steering angle is greater than a second predetermined steering angle, or if the vehicle's steering angular velocity is greater than the second predetermined steering angular velocity, then the lateral driving intention is determined to be a second lateral driving intention; and if the lateral driving intention is neither the first lateral driving intention nor the second lateral driving intention, then the lateral driving intention is determined to be a third lateral driving intention, wherein the first lateral driving intention is the driver's intention to smoothly change the route, and the second lateral driving intention is the driver's intention to suddenly change the route.

6. The method according to claim 5, wherein, The step of determining the lateral driving intention further includes: if the lateral driving intention is determined to be the first lateral driving intention, then determining whether the yaw rate of the vehicle is opposite to the sign of the steering angle of the vehicle; and if the yaw rate of the vehicle is opposite to the sign of the steering angle of the vehicle, then determining the lateral driving intention to be the fourth lateral driving intention.

7. The method according to claim 6, wherein, The step of determining the lateral driving intention further includes: if the lateral driving intention is determined to be the second lateral driving intention, then determining whether the vehicle's steering angle is opposite in sign to the vehicle's steering angular velocity; and if the vehicle's steering angle is opposite in sign to the vehicle's steering angular velocity, then determining the lateral driving intention to be the fifth lateral driving intention.

8. The method according to claim 7, wherein, The step of generating the speed distribution includes: generating an initial speed distribution based on the longitudinal driving intention; and adjusting the initial speed distribution based on the lateral driving intention.

9. The method according to claim 8, wherein, The speed distribution includes an acceleration range, a deceleration range, and a constant speed range, and the step of generating the speed distribution based on the longitudinal driving intention includes: determining the lengths of the acceleration range, the deceleration range, and the constant speed range based on the longitudinal driving intention.

10. The method according to claim 9, wherein, The step of adjusting the initial speed distribution based on the lateral driving intention includes: adjusting the lengths of the acceleration range and the deceleration range according to the lateral driving intention.

11. The method according to claim 7, wherein, The step of generating the curvature distribution includes: generating an initial curvature distribution based on the lateral driving intention; and adjusting the initial curvature distribution based on the longitudinal driving intention.

12. The method according to claim 11, wherein, If the lateral driving intention is determined to be the first lateral driving intention, then the initial curvature distribution is in the form of a linear function; if the lateral driving intention is determined to be the second or third lateral driving intention, then the initial curvature distribution is in the form of a constant function; if the lateral driving intention is determined to be the fourth lateral driving intention, then the initial curvature distribution is in the form of a quadratic function; if the lateral driving intention is determined to be the fifth lateral driving intention, then the initial curvature distribution is in the form of a step function.

13. The method according to claim 12, wherein, The step of adjusting the curvature distribution based on the longitudinal driving intention includes: multiplying the curvature distribution by a correction factor, wherein the correction factor is a value between 0 and 1, and is determined based on the longitudinal driving intention.

14. An apparatus for predicting vehicle paths, comprising: At least one memory is configured to store instructions; and at least one processor, wherein the at least one processor executes the instructions to cause the at least one processor to perform: acquiring vehicle driving information, wherein the driving information includes at least one of the vehicle's speed, acceleration, steering angle, steering angular velocity, heading angle, yaw rate, accelerator pedal and brake pedal depressure amount or gear setting; determining the driver's longitudinal driving intention based on the driving information; determining the driver's lateral driving intention based on the driving information; generating a vehicle speed distribution based on the longitudinal driving intention and the lateral driving intention; generating a vehicle curvature distribution based on the longitudinal driving intention and the lateral driving intention; and determining a predicted path for the vehicle based on the speed distribution and the curvature distribution.

15. A vehicle comprising the device of claim 14.

16. An autonomous vehicle comprising the apparatus of claim 14.

17. A non-transitory computer-readable medium comprising program instructions executable by a processor, the computer-readable medium comprising: The program instructions for acquiring vehicle driving information, wherein the driving information includes at least one of the following: vehicle speed, acceleration, steering angle, steering angular velocity, heading angle, yaw rate, accelerator pedal and brake pedal depressure amount, or gear setting; program instructions for determining the driver's longitudinal driving intention based on the driving information; program instructions for determining the driver's lateral driving intention based on the driving information; program instructions for generating the vehicle's speed distribution based on the longitudinal and lateral driving intentions; program instructions for generating the vehicle's curvature distribution based on the longitudinal and lateral driving intentions; and program instructions for determining the vehicle's predicted path based on the speed distribution and the curvature distribution.

Images