Method and Device for the Automated Driving Mode of a Vehicle, and Vehicle

Abstract

A method for autonomously driving a vehicle. When a bend is detected ahead of the vehicle, it is determined whether the visibility of a detection unit is below a threshold value. If so, the vehicle is automatically maneuvered to change lanes to an outside lane of the bend.

Claims

1-10. (canceled)

11. A method for autonomously driving a vehicle, comprising: determining whether a visibility of a detection unit is below a threshold value when a bend is detected ahead of the vehicle; and in response to determining that the visibility is below the threshold value, automatically maneuvering the vehicle to change lanes to an outside lane of the bend.

12. The method of 11, wherein the threshold value varies as a function of a current driving speed of the vehicle.

13. The method of claim 11, further comprising: determining the outside lane from map data and/or image data from a camera of the vehicle.

14. The method of claim 11, wherein the lane change is performed as a function of a traffic density detected ahead of the vehicle.

15. The method of claim 11, wherein a current speed of the vehicle is automatically adjusted based on the visibility of the detection unit.

16. The method of claim 11, wherein a required visibility of the detection unit is determined based on a predicted current braking distance of the vehicle.

17. The method of claim 11, wherein the outside lane is for travelling in the same direction as the vehicle.

18. The method of claim 11, wherein the outside lane is on a left-hand side of the vehicle in the case of a right-hand bend, and wherein the outside lane is on a right-hand side of the vehicle in the case of a left-hand bend.

19. A system for autonomously driving a vehicle, comprising: a detection unit; a computing unit, configured to: determine whether a visibility of a detection unit is below a threshold value when a bend is detected ahead of the vehicle; and in response to determining that the visibility is below the threshold value, automatically maneuver the vehicle to change lanes to an outside lane of the bend.

20. An autonomous vehicle, comprising: a detection unit; a computing unit, configured to: determine whether a visibility of a detection unit is below a threshold value when a bend is detected ahead of the vehicle; and in response to determining that the visibility is below the threshold value, automatically maneuver the vehicle to change lanes to an outside lane of the bend.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0019] FIG. 1 schematically shows an exemplary situation in which there is a route section having three lanes and a vehicle driving in an outside lane of a bend;

[0020] FIG. 2 schematically shows an exemplary situation in which there is a route section and a vehicle driving in an inner lane of a bend;

[0021] FIG. 3 schematically shows an exemplary situation in which there is a route section with a downhill gradient or incline and a vehicle in an outer lane of a bend;

[0022] FIG. 4 schematically shows an exemplary situation in which there is a route section with the downhill gradient or incline and a vehicle driving in an inner lane of a bend;

[0023] FIG. 5 schematically shows an exemplary vehicle with a detection unit and an object, located in the detection range, which object cannot be driven over by the vehicle;

[0024] FIG. 6 schematically shows an exemplary vehicle with a detection unit, a position determining unit, and a computing unit; and

[0025] FIG. 7 schematically shows an exemplary computing unit with its modules.

DETAILED DESCRIPTION OF THE DRAWINGS

[0026] Exemplary embodiments of the invention will be explained in more detail in the following with reference to the drawings. Parts that correspond to each other are provided with the same reference signs throughout the FIGS.

[0027] A route section F with three lanes F1 to F3 is shown in each of FIGS. 1 and 2, wherein the route section F is curved, i.e. has a bend K.

[0028] A vehicle 1, which is configured as a truck and is driving in automated mode, in particular without there being a vehicle user in the vehicle 1, is driving in an outside lane of the bend F1 in FIG. 1 and in an inside lane of the bend F2 in FIG. 2, wherein a central lane F3 runs between the outside lane of the bend F1 and the inside lane of the bend F2.

[0029] The route section F is in particular part of a motorway on which a large number of such self-driving vehicles 1 will be found in the future.

[0030] The vehicle 1 comprises a computing unit 2, shown by way of example in FIGS. 6 and 7, which is connected to a number of detection units 3 of a surroundings sensor system of the vehicle 1, wherein the design of the detection units 3 is based on a radar, lidar and/or camera.

[0031] In addition, the vehicle 1 has a satellite-assisted position determining unit 4, which continuously receives a position signal which is used to determine a current position of the vehicle 1.

[0032] Such a self-driving vehicle 1 is located within an existing infrastructure using detected signals of the surroundings sensor system, the position signal and also map data of a digital map C stored in the vehicle, and the driving behaviour of the vehicle 1 is adjusted in relation to road users surveyed using detected signals of the surroundings sensor system.

[0033] The surroundings sensor system installed in the vehicle has surveying characteristics which are defined by a sensor type, a design and physical boundary conditions. Typically, the surroundings sensor system represents a compromise of various functional tasks. For example, a traffic-relevant region in front of the vehicle 1 is surveyed three-dimensionally by means of a lidar-based detection unit 3, whereby semantics of a scene detected in front of the vehicle 1 are determined using detected signals of a camera-based detection unit 3, whereby road signs and light signalling systems are detected.

[0034] Requirements derived therefrom determine parameters of the respective detection unit 3, such as a base width, a focal length, an aperture angle, a pixel density, a sensor type, in particular with regard to whether signals of the camera-based detection unit 3 are detected in multiple colours or a single colour.

[0035] A method for the automated driving mode of the vehicle 1 is described hereinbelow, wherein the method focuses on a lidar-based or camera-based detection unit 3, the detection range E of which, also termed field of view or frustum, is directed in front of the vehicle 1 and the detection unit 3 is a so-called long-distance sensor.

[0036] There is no one driving ahead of the vehicle 1 and there is a high requirement for a visibility of the detection unit 3 and a requirement that a comparatively small object 5 that cannot be driven over and is shown, by way of example, in FIG. 5 is detected in order to be able to react appropriately to it. In order to avoid, as far as possible, the vehicle 1 colliding with the detected object 5, emergency braking is initiated and/or an evasive trajectory is determined, for example.

[0037] This detection unit 3 which can see a comparatively long way and is able to detect an object 5 that cannot be driven over is, as described above, a lidar-based sensor or camera sensor with a given aperture angle, wherein the detection unit 3 can also consist of multiple individual sensors.

[0038] The objective for an automated driving mode of a vehicle 1, in particular of a truck for transporting goods, is to drive at a maximum possible permissible driving speed in order, for economic reasons, to minimise a period of time during which the vehicle 1 is on the road with its load.

[0039] When taking a bend K, a visibility of the detection unit 3 directed in front of the vehicle 1 may be limited by a crash barrier, by structures and/or by plants. Such a situation applies in particular to the inside lane of the bend F2.

[0040] In order to be able to react appropriately to an object 5 that potentially cannot be driven over in the respective lane F1 to F3 by braking and/or swerving, it is necessary for the vehicle 1 to reduce its current driving speed, which increases the time period in which the vehicle 1 is in driving mode.

[0041] If, as is shown in FIG. 2, the vehicle 1 is driving in the inside lane of the bend F2, then a field of view and thus the detection range E of the detection unit 3 is restricted. If, as is shown in FIG. 1 by contrast, the vehicle 1 is driving in the outside lane of the bend F2, then the visibility is increased.

[0042] A first designation K1 depicted in FIGS. 1 and 2 is used to mark a required visibility, which is calculated from a braking distance of the vehicle 1 and is dependent on the maximum inherent deceleration of the vehicle 1 and its current driving speed.

[0043] A second designation K2 is used to show an actual visibility S of the detection unit 3, which is considerably less in FIG. 2 than in FIG. 1 if the vehicle 1 is driving in the outside lane of the bend F1. If the vehicle 1 is driving in the outside lane of the bend F1, it finds itself in the detection range E of the detection unit 3, so that the outside lane of the bend F1 is covered by sensors, in particular in relation to an object 5 that cannot be driven over.

[0044] The first designation K1, i.e. the required visibility, is located in an area B that is not visible to the detection unit 3, as shown in FIG. 2. The area B that is not visible to the detection unit 3 can also be referred to as the so-called inherent shadow of the bend.

[0045] FIGS. 3 and 4 each show a route section F with three lanes F1 to F3 and a bend K, wherein the bend K extends along a downhill gradient or crest, i.e. the route section F has a negative vertical curvature in the region of the bend K. The same applies, however, in the case of road sections having a positive vertical curvature, for example in a dip or on before an incline.

[0046] Due to the negative or positive vertical curvature, the surface of the route section F lies in a subsection G beneath or above the detection range E of the detection unit 3. The subsection G of the route section F is thus not visible to the detection unit 3.

[0047] In the case of a route section F having a downhill gradient or incline lying ahead, for example beyond a crest or a dip respectively, the actual visibility S, depicted by the second designation, of the detection unit 3 is only changed slightly by changing from the inside lane of the bend F2 to the outside lane of the bend F1, as is shown in FIGS. 3 and 4.

[0048] In the case of a route section F having a downhill gradient or incline, a visibility, shown using the first designation K1, required so as to be able to react appropriately to an object 5 that cannot be driven around in the respective lane F1, F2 of the vehicle 1 lies in the area B that is not visible to the detection unit 3. Unlike in FIG. 1, changing lanes to the outside lane of the bend F1 does not increase the actual visibility to the required visibility. The current driving speed of the vehicle 1 must therefore also be adapted after the lane change to the reduction in the actual visibility of the detection unit 3 induced by the bend and the crest or dip. Advantageously, the lane change to the outside lane of the bend F1 is not performed if it is not likely that the actual visibility can be increased to the required visibility by the lane change and if the increase in visibility likely to be achieved by the lane change is small in relation to a desired increase in visibility to the required visibility, in particular smaller than a predefined threshold. This avoids unhelpful lane changes.

[0049] If there is a downhill gradient in the route section F, a so-called road self-shadowing needs to be taken into account, whereby to reduce this, the detection unit 3 can be arranged at a comparatively high installation location on the vehicle 1.

[0050] In FIGS. 1 to 4, a three-lane route section F is selected without restriction of generality, whereby a hard shoulder is not included for reasons of simplification. In terms of visibility, the hard shoulder can be considered as a separate lane (not shown), so that statements for a two-lane route section F with a hard shoulder in this context are identical to those for the three-lane route section F shown in FIGS. 1 to 4.

[0051] FIG. 5 shows a side view of the vehicle 1 with an object 5 that cannot be driven over located in the detection range E of the detection unit 3 on the corresponding lane F1 to F3 of the vehicle 1, which object may be a load that has been lost from a vehicle (not shown) travelling in front.

[0052] FIG. 6 shows an enlarged excerpt of the vehicle 1 with the computing unit 2, the detection unit 3 and the position determining unit 4.

[0053] By way of example, FIG. 7 shows the computing unit 2 with individual modules.

[0054] According to the exemplary embodiment in FIG. 7, the computing unit 2 comprises a speed optimisation module 6, a behaviour planning module 7, a first sensor processing module SV1, a second sensor processing module SV2, a fusion module 8 and the digital map C. The behaviour planning module 7 has a situation analysis and planning module 9 and a trajectory generator 10 which is connected to an actuation system A to control steering, a drivetrain and a brake device.

[0055] A first sensor processing module SV1 processes detected signals of other sensors 11, in particular the surroundings sensor system of the vehicle 1, wherein a second sensor processing module SV2 processes detected signals of the detection unit 3.

[0056] The processed signals are then fused in a fusion module 8, wherein the speed optimisation module 6 obtains information on a traffic density prevailing on the route section F from the fusion.

[0057] A position of the vehicle 1 determined by means of the position determining unit 4 and the digital map C is supplied to the situation analysis and planning module 9.

[0058] An algorithm within the speed optimisation module 6 provides that it is determined in a first step S1 using map data from the digital map C whether a crest that cannot be seen is lying ahead of the vehicle 1. It is also determined using the digital map C what lane F1 to F3 the vehicle 1 is in. Decisions to be taken into account by the algorithm are depicted with y for yes and n for no.

[0059] If it is determined that there is no crest lying ahead of the vehicle 1, it is determined in a second step S2 whether, at a current driving speed of the vehicle 1, a bend K lying ahead of the vehicle 1 is sufficiently visible on the route section F. It is determined here in particular whether the actual visibility S of the detection unit 3 falls below a predefined threshold value, with the threshold value varying as a function of the current driving speed of the vehicle 1.

[0060] If this is not the case, it is determined in a third step S3 how dense the traffic is, wherein if the latter is determined to be not very dense, it is checked in a fourth step S4 whether the vehicle 1 is in the outside lane of the bend F1.

[0061] If the vehicle 1 is not in the outside lane of the bend F1, a lane change to the outside lane of the bend F1 is initiated in a fifth step S5.

[0062] If it is determined in the first step S1 that there is a crest lying ahead of the vehicle 1 that is not visible, or if it is determined in the third step S3 that the traffic is comparatively dense, or if it is determined in the fourth step S4 of the algorithm that the vehicle 1 is already in the outside lane of the bend F1, then the current driving speed of the vehicle 1 is adapted to the bend-induced or crest-induced visibility restriction in a sixth step S6.

[0063] If it is determined in the second step S2 that the next bend K in the route section F is sufficiently visible in accordance with the current driving speed of the vehicle 1, the driving speed is not adapted, and so the vehicle 1 continues in its automated driving mode at its current driving speed.

[0064] If, according to the fifth method step S5, a lane change of the vehicle 1 to the outside lane of the bend F1 is initiated, or if it is necessary to adapt the current driving speed in accordance with the sixth step S6, or if the driving speed does not need to be adapted, information is transmitted to the situation analysis and planning module 9 which forwards this information to the trajectory generator 10 and a trajectory applicable to a current situation is determined and sent the actuation system A.

[0065] By applying the method, the automated vehicle 1, in particular a truck for transporting goods, can be operated in a more economically optimised manner, in that comparatively unnecessary braking and re-acceleration cycles of the vehicle 1, which may be induced by cornering, are avoided as far as possible.

[0066] An average speed of the vehicle 1 can be optimised, whereby an amount of fuel and/or electrical energy consumed can also be optimised by means of a relatively steady driving mode.