Method and device for at least assisted lateral guidance of a motor vehicle

Abstract

Method for at least assisted lateral guidance of a motor vehicle, using at least one camera (2), an evaluation and control unit (3), and an actuator (4) for implementing steering interventions, wherein the evaluation and control unit (3) determines at least one lane from the data (D) of the camera (2) and generates control commands for the actuator (4) depending on the determined lane, wherein the evaluation and control unit (3) additionally takes swarm data (SD) into account, wherein in driving situations in which insufficient lane detection using the data (D) of the camera (2) is present or is to be expected, lateral guidance is carried out using the swarm data (SD) if the swarm data (SD) has previously exceeded a predetermined confidence level, wherein the confidence level of the swarm data (SD) is determined in advance by at least one comparison of the swarm data (SD) with the data (D) of the at least one camera (2), wherein a curve (Y1,Y2) for a lane boundary are determined based on the data (D) of the at least one camera (2) and the swarm data (SD) for a given distance (ΔX), wherein the area between the curves (Y1, Y2) is determined in absolute terms and compared with a threshold value S1 and / or a maximum distance (ΔYmax) between the two curves (Y1, Y2) along the distance (ΔX) is determined, wherein the maximum distance (ΔYmax) is compared with a threshold value S2.

Description

[0001] The invention relates to a method and a device for at least assisted lateral guidance of a motor vehicle.

[0002] From DE 10 2018 207 869 A1, a method for providing an automation function for a vehicle is known, in which environmental data is acquired. Depending on the acquired environmental data, the automation function is activated and a quality measure is determined. Based on this quality measure, a quality level on a multi-level scale is determined, and a graphical output is generated and displayed. This output includes a quality indicator element that is generated based on the quality measure. The environmental data can be acquired using the vehicle's environmental sensors and / or received via a data connection. The environmental data preferably relates to a lane, a lane marking, a road edge, a traffic sign, and / or another vehicle in the vicinity of the vehicle. The automation function performs at least partially autonomous lateral and / or longitudinal control of the vehicle.

[0003] German patent DE 10 2015 000 379 A1 discloses a method for supporting a target driver assistance system of a target vehicle on a section of road that cannot be measured by the target driver assistance system. In this method, a warning signal is received from a third-party driver assistance system of at least one third-party vehicle, indicating that the section of road cannot be measured by the third-party vehicle's autonomous or semi-autonomous operating mode. Furthermore, a reference signal is received from the third-party vehicle, describing its current geographical position and direction of travel. Based on this reference signal, the target vehicle marks the section of road as unmeasurable. The third-party vehicle then receives a request signal from the target driver assistance system, which includes the target vehicle's geographical position and, if applicable, its direction of travel and / or route.Depending on the request signal, it is determined whether the target vehicle is located within a predetermined area of ​​the marked track section. If so, an operating strategy is determined for the target vehicle by an external control unit, control signals are generated and transmitted to the target vehicle, where the operating strategy is then implemented.

[0004] The invention addresses the technical problem of providing a method for at least assisted lateral guidance of a motor vehicle that enables increased availability. A further technical problem is the creation of a suitable device.

[0005] The solution to the technical problem is achieved by a method having the features of claim 1 and a device having the features of claim 6. Further advantageous embodiments of the invention are set forth in the dependent claims.

[0006] The method for at least assisted lateral guidance of a motor vehicle is implemented using at least one camera, an evaluation and control unit, and an actuator for implementing steering interventions. The evaluation and control unit uses the data from the at least one camera to determine at least one lane and generates control commands for the actuator based on the determined lane. The evaluation and control unit also considers swarm data, which can be used to support the camera data. This swarm data can, for example, be in the form of a swarm map and is preferably downloaded in advance from a server or cloud for upcoming road sections. This support can consist of simple verification or plausibility checks.In driving situations where lane detection via camera data is insufficient or anticipated, lateral guidance is provided by swarm data, provided the swarm data has previously exceeded a predefined confidence level. In extreme cases, lateral guidance is based solely on swarm data. This increases the availability of lateral guidance. Lateral guidance can also be fully automated, meaning the entire process is handled by the evaluation and control unit or the actuators. Typical driving situations where insufficient lane detection via camera can occur include highway exits or motorway junctions. This is due to frequently occurring inclines or declines, as well as obscurations or even missing markings. Such a driving situation can be recognized or anticipated because, for example, navigation data indicates that a highway exit is to be used.

[0007] Additionally or alternatively, this can be achieved by driving on a deceleration lane. If the swarm data then exceeds a predefined confidence level, lateral guidance can be maintained using this data. The confidence level also serves as a measure of the quality of the swarm data. This confidence level can be binary (trustworthy or untrustworthy) or multi-level.

[0008] In one embodiment, a takeover prompt is generated for the driver if the swarm data falls below the predefined confidence level. The takeover prompt can be audible, haptic, and / or visual. Preferably, the takeover prompt is at least audible.

[0009] In another embodiment, the system detects whether the driver is holding or gripping a steering handle, generating a warning signal if lateral guidance is provided via swarm data and no re-gripping of the steering handle is detected. The detection of the steering handle being held can be achieved, for example, using capacitive or pressure-sensitive sensors in the steering handle and / or a vehicle occupant camera. The underlying principle is that, during lateral guidance based on swarm data, there may be situations where the driver needs to intervene manually. If it is detected that the driver is not holding the steering handle at all and is only lightly resting it, the warning signal alerts the driver to the possibility of needing to take over lateral guidance very quickly, thus reducing the driver's reaction time.

[0010] In another embodiment, the confidence level of the swarm data is determined by at least one comparison of the swarm data with the data from at least one camera. The advantage is that the data from the at least one camera are generally very reliable and accurate. This determination can be carried out periodically or situationally (for example, 1 km or another predetermined distance before a planned exit from the motorway).

[0011] In another embodiment, a curve for a lane boundary is determined based on the data from at least one camera and the swarm data for a predetermined stretch of road. The area between the curves is then calculated and compared to a threshold value. "Calculated" here means that areas are considered both when the curve from the swarm data lies to the left of the curve from the camera data and when it lies to the right. All areas are added together. Alternatively, in addition to the lane boundary, a curve for a lane center can also be determined and compared. Another alternative is to determine a maximum distance between the two curves along the stretch of road and compare this maximum distance to a threshold value. If this maximum distance is greater than the threshold value (e.g.,If the distance is 80 cm, it cannot be ruled out that such a deviation also exists in data from the motorway exit area or a similar traffic situation, which, although possibly isolated, could nevertheless be critical. Therefore, both conditions should preferably be checked.

[0012] Regarding the device's design, full reference is made to the preceding statements.

[0013] The invention is explained in more detail below with reference to a preferred embodiment. The figures show: Fig. 1 a schematic block diagram of a device for at least assisted lateral guidance of a motor vehicle and Fig. 2 two curves of a lane boundary using data from at least one camera and swarm data.

[0014] In the Fig. Figure 1 is a simplified representation of a device 1 for at least assisted lateral guidance of a motor vehicle, preferably enabling fully automated lateral guidance. The device 1 comprises at least one camera 2, an evaluation and control unit 3, and an actuator 4 for implementing steering interventions. The evaluation and control unit 3 is configured to determine at least one lane from the data D of the at least one camera 2 and to generate control commands for the actuator 4 based on the determined lane. The actuator 4 is a steering system of the motor vehicle. Furthermore, the device 1 comprises a receiver 6 that can receive swarm data SD from an external server or cloud and store it in a memory 7. The evaluation and control unit 3 can access the swarm data SD.Furthermore, the device 1 includes a position determination unit 8, which is designed, for example, as a satellite-based positioning system such as GPS. The evaluation and control unit 3 can also access the data from the position determination unit 8 in order to, for example, locate the swarm data SD. The device 1 also includes output means 9 by which a driver can be warned or informed visually and / or audibly and / or haptically. Finally, the device 1 includes a steering handle 10 in the form of a steering wheel, the steering handle being equipped with a rotation angle or torque sensor 11, the signals of which are supplied to the actuator 4 when the driver steers manually. Capacitive sensors 12 are arranged on the steering handle 10, by means of which hands-on detection can be performed. It can therefore be detected whether the driver is gripping the steering handle or not.

[0015] The data from sensors 12 are also fed to the evaluation and control unit 3.

[0016] Under normal operating conditions, the vehicle is automatically guided laterally based on data from at least one camera 2, whose data is typically highly accurate. Swarm data SD may be used to supplement data D, but this is not mandatory. If the evaluation and control unit 3 detects a driving situation where insufficient lane detection using camera 2 data is expected (e.g., a highway exit), a confidence level for the swarm data SD is checked. The upcoming highway exit can be determined, for example, from navigation data of a calculated route or from a turn onto a deceleration lane.

[0017] The determination and verification of the confidence level of the swarm data SD will now be briefly described using the following examples: Fig.2. This will be briefly explained. For this purpose, the evaluation and control unit 3 first determines a lane boundary Y1 from the data D of camera 2 and a lane boundary Y2 from the swarm data SD. The two curves Y1 and Y2 can also intersect. The two curves Y1 and Y2 are then compared over a predefined distance ΔX of, for example, 50 m to 100 m, where X is oriented in the direction of travel. The comparison consists, on the one hand, of determining the area between the two curves Y1 and Y2 and comparing it with a first threshold value S1. If the area is smaller than the threshold value S1, this indicates that the confidence level is high. Additionally, the maximum distance ΔY is also determined. max The area between the two curves is determined and compared with a second threshold S2. The swarm data SD is then considered reliable if the area is smaller than the first threshold S1 and the maximum distance ΔY is less than the specified value. maxThe second threshold S2 is smaller than the swarm data (confidence measure is greater than a threshold). If the swarm data SD is reliable, fully automated lateral guidance can be performed based on the swarm data SD. However, the driver will be informed, as they may need to take over lateral guidance or intervene. This information can be provided more urgently if it is determined that the driver is not currently gripping the steering handle 10.

[0018] If, however, the swarm data SD is not trustworthy, the driver will be asked to take over lateral control again.

[0019] It should be noted that the two test steps for the swarm data SD can be supplemented by further test steps, whereby the test results can then be merged differently. Reference symbol list 1 Device 2 cameras 3 Evaluation and control unit 4 Actuators 6 Receiving equipment 7 storage 8 Position determination unit 9. Expenditure funds 10 Steering handle 11 Torque sensor 12 Sensor SD swarm data D Camera data S1, S2 thresholds Y1, Y2 curves ΔX extent DY max maximum distance QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] DE 102018207869 A1

[0002] DE 102015000379 A1

[0003]

Claims

[1] Method for at least assisted lateral guidance of a motor vehicle, using at least one camera (2), an evaluation and control unit (3) and an actuator (4) for implementing steering interventions, wherein the evaluation and control unit (3) determines at least one lane from the data (D) of the camera (2) and generates control commands for the actuator (4) depending on the determined lane, wherein the evaluation and control unit (3) additionally takes swarm data (SD) into account, wherein in driving situations in which insufficient lane detection using the data (D) of the camera (2) is present or is to be expected, the lateral guidance is carried out using the swarm data (SD) if the swarm data (SD) has previously exceeded a predetermined confidence level. [2] Method according to claim 1, characterized by , that a takeover request is generated for the driver if the swarm data (SD) falls below the specified confidence level. [3] Method according to claim 1 or 2, characterized by , that the driver's holding of a steering handle (10) is detected, generating a warning signal if lateral guidance is provided by means of swarm data (SD) and no repositioning of the steering handle (10) has been detected. [4] Method according to any of the preceding claims, characterized by , that the confidence level of the swarm data (SD) is determined by at least one comparison of the swarm data (SD) with the data (D) of the at least one camera (2). [5] Method according to claim 4, characterized by , that a curve (Y1, Y2) for a lane boundary is determined based on the data (D) of the at least one camera (2) and the swarm data (SD) for a given distance (ΔX), whereby the area between the curves (Y1, Y2) is determined in magnitude and compared with a threshold value S1 and / or a maximum distance (ΔY max) between the two curves (Y1, Y2) along the path (ΔX) is determined, where the maximum distance (ΔY) max ) is compared with a threshold value S2. [6] Device (1) for at least assisted lateral guidance of a motor vehicle, comprising at least one camera (2), an evaluation and control unit (3) and an actuator (4) for implementing steering interventions, wherein the evaluation and control unit (3) is configured to determine at least one lane from the data (D) of the at least one camera (2) and to generate control commands for the actuator (4) depending on the determined lane, wherein the evaluation and control unit (3) additionally takes swarm data (SD) into account, wherein the evaluation and control unit (3) is further configured such that, in driving situations in which insufficient lane recognition by means of the data (D) of the camera (2) is present or is to be expected, the lateral guidance is carried out by means of the swarm data (SD) if the swarm data (SD) has exceeded a predetermined confidence level. [7] Device according to claim 6, characterized by, that the evaluation and control unit (3) is designed to generate a takeover request for the motor vehicle driver when the swarm data (SD) has fallen below the specified confidence level. [8] Device according to claim 6 or 7, characterized by , that the device (1) has at least one sensor for detecting a holder of a steering handle, wherein the evaluation and control unit (3) is designed to generate a warning signal when the lateral guidance is carried out using the swarm data (SD) and no re-gripping of the steering handle (10) has been detected. [9] Device according to any one of claims 6 to 8, characterized by , that the evaluation and control unit (3) is designed to determine the confidence level of the swarm data (SD) by at least one comparison of the swarm data (SD) with the data (D) of the at least one camera (2). [10] Device according to claim 9, characterized by, that the evaluation and control unit (3) is designed such that a curve (Y1, Y2) for a lane boundary is determined based on the data (D) of the at least one camera (2) and the swarm data (SD) for a predetermined distance (ΔX), wherein the area between the curves (Y1, Y2) is determined operationally and compared with a threshold value S1 and / or a maximum distance (ΔY max ) between the two curves (Y1, Y2) along the path (ΔX) is determined, where the maximum distance (ΔY) max ) is compared with a threshold value S2.

Images