A computer-implemented method for providing steering control for autonomous or semi-autonomous vehicles.

Abstract

The present invention relates to a computer-implemented method for providing steering control for an autonomous or semi-autonomous vehicle (2), comprising the following steps: acquiring road-related sensor data of the moving vehicle (2) from at least one on-board sensor (S100); determining a first target point in front of the moving vehicle (2) (S200); and determining a second target point (T2) located on the same lateral plane (P) as the first target point (T1) at a time when the vehicle (2) has traveled a certain amount of time from the time the first target point (T1) was determined, and / or after the vehicle (2) has traveled a predetermined longitudinal distance (LC), thereby determining the first target point (T1) The present invention relates to a method characterized by comprising the steps of: determining a second target point (T2) that lies on the same lateral plane (P) as the first target point (T1, T2) (S300); calculating the lateral distance (p_dp) between the determined first and second target points (T1, T2) (S400); comparing the calculated lateral distance (d_pd) with at least one predetermined first threshold (S500); determining the trajectory tracking error if the calculated lateral distance (d_pd) exceeds at least the predetermined first threshold (S600); and supplying a control signal to the steering controller depending on the determined tracking error (S700).

Description

Technical Field

[0004] , , , , , , , , , ,

[0005] , , , , , , ,

[0003]

[0001] The present invention relates to a method, a program element, and a computer-readable medium, a driver assistance system, and a vehicle equipped with the driver assistance system, which are implemented on a computer for providing steering control for an autonomous or semi-autonomous driving vehicle.

Background Art

[0002] Modern driver assistance systems assist the driver by performing specific driving functions or enable the vehicle to travel autonomously without the need for driver intervention. The autonomous driving system includes lateral vehicle control systems such as a lane centering system and a lane departure prevention system, which aim to keep the vehicle within the driving lane.

[0003] US9428187B2 discloses a system and method for providing the planning and creation of a trajectory for automatically maintaining the center of a lane for a vehicle traveling on a road. In particular, within the effective range of a camera, road measurement values from a vision camera are used, and outside the effective range of the camera, road measurement values from a map database are used. The method uses road measurement values from the camera to determine a target trajectory along a first segment of the road and identifies the end of the first segment based on how accurately the camera captures the road. The camera estimates the position and orientation of the lane relative to the center of the vehicle based on points. For example, the road is modeled as two sections of a quadratic polynomial.

Prior Art Documents

Patent Literature

[0004]

Patent Literature 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] Within the camera's field of view, the ability to detect lane trajectories may be limited due to road curves. In addition, the vehicle's own motion can introduce additional positional errors in the mathematical transformation from image to world space. Therefore, the captured or fused lane model may not provide the trajectory prediction accuracy required in autonomous or semi-autonomous driving systems. Inaccurate lane models can lead to critical driving situations. [Means for solving the problem]

[0006] Therefore, improved methods are needed to reliably guide autonomous or semi-autonomous vehicles. This challenge is addressed by the main subject of the independent claim. Embodiments are disclosed by the dependent claims, the description in the following specification, and the accompanying drawings.

[0007] It should be noted that while all embodiments of the present invention relating to the method can be carried out according to the steps described, this is by no means the only or essential order of steps. That is, unless otherwise explicitly stated below, the methods described herein can be carried out in a different order of steps without departing from the respective embodiments of the method.

[0008] According to the first aspect, a computer-implemented method for providing steering control for autonomous or semi-autonomous vehicles is provided. In this context, the term "semi-autonomous" means that one or more of the driver's driving functions are performed by the vehicle's driver assistance system, while the term "autonomous" means that all driving functions are performed by the vehicle's driver assistance system.

[0009] This method involves the following steps: First, sensor data about the road from the moving vehicle is acquired from at least one on-board sensor. For example, the on-board sensor could be a camera that captures images of the road for extracting lane parameters.

[0010] In the next step, a first target point in front of the moving vehicle is determined. Specifically, the first target point is the target position to which the vehicle should be guided. For example, the location of the first target point is on the centerline of the ego lane. Determining the first target point can be carried out depending on several factors, such as planning strategy and system delay. In particular, the longitudinal distance between the first target point and the vehicle's current position at the time the first target point is determined may depend, for example, on an absolute distance value of 60 meters and / or on a time-based value such as 2 seconds, which depends on the current vehicle speed.

[0011] According to the present invention, a second target point, which lies on the same lateral plane as the first target point, is determined when the vehicle has traveled a certain amount of time and / or a predetermined longitudinal distance from the time the first target point was determined. In this context, the term "on the same lateral plane" means that the two points do not need to coincide precisely, and tolerances, limits of measurement accuracy, and other factors well known to those skilled in the art may exist, as long as their characteristics do not hinder the intended effect. In particular, the determination of the second target point is performed when the vehicle has reached the determined first target point, or at least is near the determined first target point. The term "near" the determined first target point means that it is within a close range to the first target point, i.e., for example, the longitudinal distance between the current position of the moving vehicle and the first target point is about 1 meter or less. Sensor detection, in particular images provided by on-board cameras, is usually very accurate within a close range to the vehicle. Therefore, the identified second target point can favorably be considered as pseudo-ground truth.

[0012] The next step is to calculate the lateral distance between the identified first and second target points and compare the calculated lateral distance with at least one predetermined first threshold. If the calculated lateral distance exceeds at least the predetermined first threshold, the trajectory tracking error is determined. This provides a reliable consistency verification of the lane model.

[0013] Next, a control signal is supplied to the steering controller, depending on the identified tracking error. In particular, if a tracking error is identified, the performance of the steering control, preferably the lane keeping assist, will deteriorate.

[0014] The advantage of this lane model consistency is that it reduces the potential risk of erroneous steering behavior due to undetectable lane model misdetections resulting from sensing and fusion. The information needed to determine the accuracy of the lane model is derived in a simple and resource-efficient manner.

[0015] According to one embodiment, the determined lateral interval can be a single determined value and / or an average value calculated from multiple determined lateral interval values. For example, the lateral interval can be checked at regular intervals. Such intervals may be a predetermined calculation period, image period, or travel distance.

[0016] In another preferred embodiment, the calculated lateral distance is compared to at least a predetermined second threshold, the second threshold being set higher than the first threshold. Preferably, a first steering control signal is supplied when at least the predetermined first threshold is exceeded, and a second steering control signal is supplied when the predetermined second threshold is exceeded. Furthermore, the predetermined first and second control signals can be at least partially different from each other. In particular, the predetermined second control signal imposes stricter constraints on autonomous or semi-autonomous driving than the predetermined first control signal.

[0017] Specifically, if the calculated lateral distance continues to exceed a predetermined time and at least a predetermined first threshold, the steering wheel handoff time is shortened, and / or the driver is notified that the current steering control system may be unstable, and / or the threshold for the driver to intervene in the steering control system is lowered. By considering time thresholds in this way, simple measurement errors can be eliminated. Reducing the handoff time and / or notifying the driver in cases where inaccurate trajectory guidance is expected ensures that the driver can take over control in a timely manner.

[0018] In another embodiment, if the calculated lateral distance continues to exceed a predetermined second threshold for a longer period than a predetermined time, a handover request is sent to the driver and / or the steering control is suspended for a predetermined time. By imposing an obligation on the driver to hand over control, misguided steering control is prevented in cases where inaccurate trajectory tracking is clearly expected.

[0019] In a further embodiment, the first target point is received and stored from the lane model as the horizontal and vertical axes of a Cartesian coordinate system. In other words, the position of the first target point is measured directly as relative coordinates and later converted to absolute coordinates. This coordinate system can be transmitted from the sensing or fusion process via a mathematical model such as a polynomial model, a clothoid model, or a polyline model. Furthermore, the first target point in Cartesian coordinates can be converted to a global coordinate system as the three-dimensional position coordinates of the vehicle.

[0020] In a further embodiment, the second target point is received and stored from the lane model as the horizontal and vertical axes of a Cartesian coordinate system, and the second target point in Cartesian coordinates is transformed into the same global coordinate system as the three-dimensional position coordinates of the first target point, as the three-dimensional position coordinates of the vehicle. This provides a stable and highly accurate three-dimensional model of the road and / or the three-dimensional position coordinates of the vehicle.

[0021] In a further embodiment, the lateral distance is calculated within the Cartesian coordinates of the first and second target points and, in the next step, is transformed to obtain a three-dimensional model of the road as the three-dimensional position coordinates of the vehicle. In this way, once the first and second target points are determined, the lateral deviation between them can be accurately determined.

[0022] In another embodiment, a computer program element is provided, which, when executed by a computer, includes instructions causing the computer to perform a method based on one or more of the method steps described above.

[0023] In another embodiment, a computer-readable medium is provided. When executed by a computer, the computer-readable medium contains instructions that cause the computer to perform a method based on one or more of the method steps described above.

[0024] In yet another embodiment, a driver assistance system for a vehicle is provided. This driver assistance system is configured to control the vehicle autonomously or semi-autonomously. This means that at least part of the vehicle's guidance is performed by the vehicle itself. Specifically, this applies to vehicles with SAE (Society of Automotive Engineers) autonomous driving level 2 or higher. For example, the driver assistance system may include automatic lane centering assist, active lane change assist, adaptive cruise control, and emergency steering assist.

[0025] The driver assistance system includes an in-vehicle sensor unit for detecting the road of a vehicle in motion. The in-vehicle sensor unit can include at least one camera, radar, lidar, and / or ultrasonic sensor. Furthermore, the driver assistance system also includes a control unit that receives sensor data corresponding to the road of the vehicle in motion from at least the in-vehicle sensor unit. In particular, the control unit is configured to be able to control the speed and / or steering of the vehicle.

[0026] The control unit is configured to determine a first target point in front of the vehicle in motion. The first target point can be determined based on distance or time. Furthermore, the determination of the first target point can be carried out based on a plurality of factors such as a planning strategy and system delay. For example, when the current ego speed is 40 m / s, the first target point is specified 2 seconds ahead.

[0027] The control unit is configured to determine a second target point on the same lateral plane as the first target point when the vehicle has traveled for a certain time from the moment of determining the first target point, and / or the control unit is configured to determine a second target point on the same lateral plane as the first target point after the vehicle has traveled a predetermined distance in the longitudinal direction.

[0028] The control unit is configured to calculate the lateral distance between the identified first and second target points and to compare the calculated lateral distance with at least one predetermined first threshold. Furthermore, the control unit is configured to determine the trajectory tracking error if the calculated lateral distance exceeds at least the predetermined first threshold. The closer the distance between the vehicle and the target point, the higher the image resolution and the better the measurement accuracy of the target point. Therefore, by comparing the first and second target points, it becomes possible to efficiently verify the consistency of the lane model.

[0029] If the tracking error exceeds at least a predetermined first threshold, it becomes impossible to maintain a stable and highly accurate lane model in lane tracking. Therefore, the control unit is configured to supply control signals to the steering controller in accordance with the determined tracking error. For example, if the tracking error exceeds at least a predetermined first threshold, the steering control and / or speed control are stopped in an appropriate manner.

[0030] In particular, the control unit may include a microcontroller, processor, central processing unit (CPU), graphics processing unit (GPU), digital signal processing unit (DSP), ASIC (application-specific integrated circuit), FPGA (field-programmable gate array), and software for executing corresponding processing procedures. Therefore, the present invention can be implemented using digital electronic circuits, computer hardware, firmware, or software.

[0031] Hereinafter, exemplary embodiments of the present invention will be described with reference to the accompanying drawings. [Brief explanation of the drawing]

[0032] [Figure 1] Figure 1 shows a schematic top view of an exemplary vehicle equipped with a driver assistance system according to one embodiment of the present invention; [Figure 2]Figure 2 shows a flowchart illustrating the procedure for a computer-implemented method for providing steering control for an autonomous or semi-autonomous vehicle according to one embodiment of the present invention. [Modes for carrying out the invention]

[0033] The accompanying drawings are included to provide a further understanding of the present invention and constitute part of this specification. The drawings illustrate specific embodiments of the present invention and, together with the description, illustrate the principles of the present invention. Other embodiments of the present invention and many of the advantages associated therewith will be readily apparent by referring to the detailed description below.

[0034] Figure 1 is a top view showing a driver assistance system 1 for autonomous or semi-autonomous control of a vehicle 2 traveling in lane 3. The driver assistance system 1 includes control units that can be used to control the vehicle's driving functions or autonomous driving systems such as a lane keeping system, adaptive cruise control system, autonomous emergency braking system, construction zone driving assistance, traffic jam driving assistance, highway / driving assistance function, and autopilot function. In this embodiment, the driver assistance system 1 is a lane keeping assist adapted to guide the vehicle 2 along the center line C of lane 3.

[0035] The driver assistance system 1 includes an on-board sensor unit for detecting the road ahead of the moving vehicle 2. The sensor unit may include one or more cameras. The control unit is configured to receive sensor data corresponding to the road ahead of the moving vehicle 2 from at least the on-board sensor unit. In addition, information corresponding to the road ahead of the moving vehicle 2 can be obtained via Car-to-X communication and / or Car-to-Car communication.

[0036] The control unit is configured, as the next step, to determine a first target point T1 at time T=0 in front of the moving vehicle 2. The first, second, and subsequent determined target points are waypoints indicating the planned route of the vehicle. The first target point T1 is determined as x_previewdistance_CVS and y_previewdistance_CVS in Cartesian coordinates, which can be converted to x_previewdistance_GCS and y_previewdistance_GCS in the global coordinate system as the three-dimensional position coordinates of the vehicle 2. In particular, the first, second, and subsequent determined target points are based on the odometry data of the moving vehicle.

[0037] When the distance from the vehicle is far (for example, more than 60 meters), it is difficult to perform highly accurate lane tracking, so the consistency check of the lane model is performed as follows. The control unit is configured to determine a second target point T2 that lies on the same lateral plane P as the first target point when the vehicle 2 has traveled for a certain period of time from the time the first target point T1 was determined. That is, in this embodiment, after the time T=0 when the first target point T1 was determined, the vehicle 2 travels for a predetermined time T=t. Alternatively, or additionally, the control unit is configured to determine a second target point T2 that lies on the same lateral plane P as the first target point T1 after the vehicle 2 has traveled a predetermined distance LC in the longitudinal direction. The predetermined distance LC may be configured such that when the second target point T2 is measured, the vehicle 2 approaches the position of the first target point T1, which was determined, with an error of less than 1 meter, for example using odometry. The lateral plane P is oriented laterally with respect to the direction of the ego lane 3, and particularly preferably perpendicular to the direction of the lane.

[0038] The control unit is configured to calculate the lateral distance d_pd between the first and second target points T1 and T2, particularly preferably the lateral distance d_pd in Cartesian coordinates. The calculated lateral distance d_pd is compared with at least one predetermined first threshold. To perform image-based three-dimensional lane detection, the first and second target points T1 and T2 may be transformed into the global coordinate system GCS as three-dimensional position coordinates T1 x_pd_GCS, y_pd_GCS and T2 x_current_GCS, y_current_GCS.

[0039] The control unit is configured, as the next step, to determine a trajectory tracking error if the calculated lateral spacing exceeds at least the predetermined first threshold. According to Figure 1, the first and second target points T1 and T2 have a lateral spacing d_pd between them. Furthermore, the control unit is configured to supply a control signal to the steering controller in accordance with the determined tracking error. In this embodiment, the lateral spacing d_pd exceeds at least the predetermined first threshold. According to this embodiment, based on the results obtained, the adaptive lane centering function may be functionally limited based on a consistency check of the lane model in accordance with the determined tracking error. That is, advantageously, control for autonomous or semi-autonomous vehicles is provided only when lane tracking is stable and highly accurate. However, in other exemplary embodiments, the determined lateral spacing d_pd may be compared to at least one predetermined second threshold that is stricter than the predetermined first threshold. Specifically, if the lateral spacing d_pd exceeds a predetermined first threshold, automatic or semi-automatic control is restricted, and if the lateral spacing d_pd exceeds a predetermined second threshold, automatic or semi-automatic control can be stopped. If the calculated lateral spacing d_pd falls below the first threshold, the lane model can be used again for the automatic driving function.

[0040] Figure 2 shows a flowchart diagram for a computer-implemented method for providing steering control for an autonomous or semi-autonomous vehicle according to a certain embodiment. The method includes, as a first step, step S100, receiving sensor data from at least one on-board sensor corresponding to the road on which the vehicle is traveling. A first target point T1 is determined in front of the traveling vehicle 2 in step S200, and then a second target point T2, which lies on the same lateral plane P as the first target point T1, is determined in step S300 when the vehicle 2 has traveled a certain amount of time from the time the first target point T1 was determined, and / or after the vehicle 2 has traveled a predetermined distance LC in the longitudinal direction. Particularly preferably, the second target point is determined within an area close to the vehicle 2. In the next step S400, the lateral distance between the determined first and second target points is calculated, and then the calculated lateral distance d_pd is compared with at least one predetermined first threshold in step S500. If the calculated lateral spacing exceeds at least the predetermined first threshold, a trajectory tracking error is determined in step S600. In the next step S700, a control signal is supplied to the steering controller depending on the determined tracking error. Those skilled in the art will readily recognize that the lane model consistency check step can be performed only once or multiple times. In another embodiment, the lateral spacing d_pd can also be calculated when collecting target points T1, T2 which are determined iteratively or averaged.

[0041] Further modifications of embodiments of the present disclosure can be understood and implemented by a person skilled in the art in the course of implementing the present invention by examining the drawings and the present disclosure and the appended claims. In the claims, the term “comprising” (including, comprising, consisting of) is not exclusive of other elements or steps, and the indefinite article “a” or “an” is not exclusive of plural. A single processor or other unit may also perform the functions of multiple items or steps described in the claims. Where certain means are described in different dependent claims, this does not mean that those means cannot be combined and used advantageously. Computer programs may be stored in or distributed on suitable media such as optical storage media or semiconductor storage media provided together with or as part of other hardware, or they may be distributed in other forms, such as via the Internet or other wired or wireless communication systems. Reference numerals in the claims should not be construed as limiting the scope of the claims.

Claims

1. A computer-implemented method for providing steering control for autonomous or semi-autonomous vehicles (2): Step (S100): Obtain sensor data related to the road from at least one on-board sensor of a moving vehicle (2), Step (S200) to determine the first target point (T1) in front of the moving vehicle (2), Step (S300) to determine the second target point (T2) which lies on the same lateral plane (P) as the first target point (T1) when the vehicle (2) has traveled a certain amount of time from the time the first target point (T1) was determined, and / or after the vehicle (2) has traveled a predetermined distance (LC) in the longitudinal direction. Step (S400) to calculate the lateral distance (p_dp) between the determined first and second target points (T1, T2), Step (S500): Compare the calculated lateral interval (d_pd) with at least one predetermined first threshold. If the calculated lateral spacing (d_pd) exceeds at least the predetermined first threshold, the trajectory tracking error is determined in step (S600). Step (S700): A control signal is supplied to the steering controller based on the calculated tracking error. A method characterized by encompassing the following.

2. The method according to claim 1, wherein the calculated lateral gap (d_pd) is compared with at least a predetermined second threshold, wherein the second threshold is set higher than the first threshold, wherein a first steering control signal is supplied when it exceeds at least the predetermined first threshold, and a second steering control signal is supplied when it exceeds the predetermined second threshold, wherein the predetermined first and second control signals are at least partially different from each other.

3. The method according to 1 or 2, characterized in that if the calculated lateral distance (d_pd) continues to exceed a predetermined time and at least a predetermined first threshold, the steering wheel handoff time is shortened and / or the driver is informed that the current steering control system may be unstable and / or the threshold for the driver to intervene in the steering control system is lowered.

4. The method according to 2 or 3, characterized in that if the calculated lateral distance (d_pd) is longer than a predetermined time and continues to exceed a predetermined second threshold, a handover request is sent to the driver and / or steering control is suspended for a predetermined time.

5. The method according to any one of the prior claims, characterized in that a first target point (T1) is received and stored from a lane model as the horizontal and vertical axes (x_pd_CVS, y_pd_CVS) of a Cartesian coordinate system, and the first target point (T1) in Cartesian coordinates is converted to a global coordinate system as the three-dimensional position coordinates (x_pd_GCS, y_pd_GCS) of a vehicle (2).

6. The method according to 5, characterized in that a second target point (T2) is received and stored from the lane model as the horizontal and vertical axes (x_cur_CVS, y_cur_CVS) of a Cartesian coordinate system, and the second target point (T2) in Cartesian coordinates is transformed into the same global coordinate system as the three-dimensional position coordinates of the first target point (T1) as the three-dimensional position coordinates of the vehicle (2).

7. The method according to 5 or 6, characterized in that the lateral spacing (d_pd) is calculated in the Cartesian coordinates of the first and second target points (T1, T2) and, in the next step, is transformed to obtain a three-dimensional model of the road as the three-dimensional position coordinates of the vehicle.

8. A computer program element wherein, when executed by a computer, the program element contains instructions that cause the computer to perform the method according to any one of claims 1 to 7.

9. A computer-readable medium that, when executed by a computer, contains instructions causing the computer to perform the method described in any one of claims 1 to 7.

10. An on-board sensor unit for detecting the road surface of a moving vehicle, A control unit that receives sensor data corresponding to the road conditions of a moving vehicle from an onboard sensor unit. A driver assistance system for the vehicle (2) that includes: The control unit is configured to determine the first target point in front of the moving vehicle. The control unit is configured to determine a second target point (T2) that lies on the same lateral plane as the first target point when the vehicle (2) has traveled a certain amount of time from the time the first target point (T1) was determined, and / or to determine a second target point (T2) that lies on the same lateral plane (P) as the first target point (T1) after the vehicle (2) has traveled a predetermined distance (LC) in the longitudinal direction. The control unit is configured to calculate the lateral distance between the first and second target points that have been determined, and to compare the calculated lateral distance with at least one predetermined first threshold. The control unit is configured to determine the trajectory tracking error when the calculated lateral distance exceeds at least the predetermined first threshold. The control unit is configured to supply control signals to the steering controller in accordance with the identified tracking error. A driver assistance system (1) characterized by the following.

11. A vehicle (2) comprising the driver assistance system (1) according to claim 10.

Images