Method and control device for adapting the driving behavior of a vehicle by means of a control device

Abstract

A method for matching the driving behavior of a vehicle by means of a control device is disclosed, wherein measurement data collected by at least one radar sensor and / or lidar sensor are received and at least one obstacle within the scanning region of the radar sensor and / or lidar sensor is ascertained by analyzing the received measurement data, a section of the scanning region that is obscured by the obstacle is ascertained on the basis of the ascertained obstacle and the installation position of the radar sensor and / or lidar sensor, an object prediction is created for the obscured section of the scanning region that a dynamic object can potentially cross the lane of the vehicle from the obscured section of the scanning region, and the driving behavior of the vehicle is matched depending on the situation on the basis of the object prediction. Furthermore, a control device, a computer program and a machine-readable storage medium are disclosed.

Description

Technical Field

[0001] The invention relates to a method for adapting the driving behavior of a vehicle. The invention also relates to a control device, a computer program, and a machine-readable storage medium. Background Art

[0002] Radar sensors, by exploiting the Doppler effect, can measure the relative radial velocity between the radar sensor and the reflecting object. Furthermore, radar sensors can be used to determine the position of an object. For these reasons, radar sensors are widely used in driver assistance functions for vehicles. Radar sensors can be used in vehicles of all automation levels, for example, to avoid collisions with dynamic objects or at least generate collision warnings. To this end, the radar sensor's scanning area, or detection zone, is scanned to detect both static and dynamic objects.

[0003] Because the measurement methodology is based on the Doppler effect, the radar beam emitted to study the traffic environment can be blocked by static obstructions or obstacles. Due to the obstruction, a section of the scanning area can no longer be scanned by the radar beam. Such obstacles can be dynamic or static objects that block the propagation of the radar beam. In addition, weather-related influences such as thick fog or smoke can also prevent the radar beam from propagating along the scanning area. As a result, such obstacles lead to the loss of detection of dynamic objects that could potentially cause a collision. The probability of a collision is particularly increased if the dynamic object is located behind an obstruction and can only be detected by the sensor after it has escaped from the obscured section. Summary of the Invention

[0004] The object of the present invention can be seen as to be to provide a method for increasing traffic safety on shaded sections or obstacles.

[0005] This object is achieved by means of the method, control device, computer program, and machine-readable storage medium described in accordance with the present invention. Advantageous embodiments of the present invention are explained below.

[0006] According to one aspect of the present invention, a method is provided for adapting the driving behavior of a vehicle by means of a control device.

[0007] In one step, measurement data collected by at least one radar sensor and / or lidar sensor is received. The received measurement data is evaluated to determine at least one obstacle within the scanning range of the radar sensor and / or lidar sensor. The obstacle can be a dynamic or static object that is at least temporarily parked.

[0008] Obstacles can be formed, for example, by construction sites, parked vehicles, parked trailers or containers, vegetation, buildings, smoke or fog, rain or snow, and the like.

[0009] Thus, in a first step, obstacles can be detected in terms of their position and geometry. This detection can be done, for example, by comparing measurement data with a radar-based map. Alternatively or additionally, obstacles or obstructions to the radar and / or laser beam can be detected using optical sensors based on the position and geometry of the object.

[0010] In a further step, the section of the scanning area obscured by the obstacle is determined based on the determined obstacle and the installation position of the radar sensor and / or lidar sensor. The obscured area or section can thus be determined based on the sensor's installation position on the vehicle and the linearly propagating radar or laser beam.

[0011] An object prediction is then generated for the obscured section of the scanning area. This object prediction takes into account that dynamic objects can potentially cross the vehicle's lane from the obscured section of the scanning area. If the dimensions of the obscured section are known, it can be assumed that a dynamic object located in the obscured section can escape from it at any time and cross the vehicle's lane. Once the object is in view or is no longer obscured by an obstacle, it is visible to the vehicle's sensors.

[0012] This method does not explicitly determine dynamic objects behind obstacles; instead, it assumes the risk of a dynamic object escaping from the obstructed section. This risk is taken into account by correspondingly adapting the driving behavior. The vehicle's driving behavior is adapted situationally and based on object predictions. The object assumption or object prediction can be implemented, for example, as a probability that a dynamic object will cross the vehicle's roadway.

[0013] Since the probability of avoiding a collision depends on various conditions, for example on the relative object distance and on the total braking distance that can be achieved, the probability of a collision increases in the case of a vehicle approaching a section that is obscured. By means of the method it is possible to implement a situation-adapted vehicle behavior in the vicinity of obstacles and obscured areas in order to avoid potential collisions. By predicting dynamic objects in the obscured section of the scanning area, a matched vehicle behavior can be defined such that an increased traffic safety is ensured when driving through the obscured section of the scanning area.

[0014] According to another aspect of the application, a control device is provided, wherein the control device is configured to implement the method. The control device can be, for example, a vehicle-side control device, a control device outside the vehicle or a server unit outside the vehicle, for example a cloud system.

[0015] Furthermore, according to an aspect of the application, a computer program is provided, which comprises instructions for causing a computer or a control device to implement the method according to the application when said computer program is executed by the computer or the control device. According to another aspect of the application, a machine-readable storage medium is provided, on which the computer program according to the application is stored.

[0016] The vehicle can be operated according to the German Federal Highway Bureau (BASt) standards assisted, partially automated, highly automated and / or fully automated, that is to say driverless.

[0017] The vehicle can be, for example, a passenger vehicle, a goods vehicle, a robot taxi or the like. The vehicle is not limited to driving on roads. Rather, the vehicle can also be configured as a water-borne vehicle, an air-borne vehicle, for example a transport drone, and the like.

[0018] The method can be used particularly advantageously in vehicles that can be operated highly automated and fully automated. Here, in particular, obstacles can be detected in intersection scenarios with vehicles or objects parked at the roadside in urban environments and can be taken into account by means of a situation-adapted driving behavior. Furthermore, the method can also be used for overtaking maneuvers, since a dynamic object that is obscured can be located behind an object or vehicle to be overtaken.

[0019] In an embodiment, the method is executed as part of a collision prediction, or the detected objects and the situation-adapted driving behavior of the vehicle are used as input variables for a collision prediction. By means of the measures, the method can be integrated particularly effectively into existing control devices and control modules of vehicles.

[0020] The method can be implemented as a hardware-side and / or software-side module that can influence the driving behavior of the vehicle. Here, the actuators of the vehicle can be controlled by the method to achieve longitudinal guidance and / or transverse guidance of the vehicle.

[0021] Based on this method, in particular, conclusions can be made about visibility: in which ranges of the theoretical scanning range or detection range of the radar sensor no reliable radar data can be generated. Based on this knowledge, object hypotheses are derived about potential dynamic objects.

[0022] According to another embodiment, data from a digital radar map is received. During the evaluation of measurement data collected by the radar sensor, the collected measurement data is compared with data from the digital radar map to determine obstacles and sections of the scanning area obscured by the obstacles. This allows for the ascertainment of static objects and obstacles, wherein a previously recorded radar map depicting the unobstructed vehicle surroundings can be used. Comparing the radar map with the ascertained measurement data during driving allows obstacles to be identified. If the radar map contains certain locations or reflection points that cannot be measured during driving, sections obscured by obstacles can be expected.

[0023] The at least one obstacle can be a static or dynamic object. Furthermore, the obstacle can be caused by atmospheric or weather-related influences, such as fog, rain, smoke, or the like. The obstacle blocks the propagation of the electromagnetic beam at least regionally and thus shields a section of the scanning area.

[0024] According to another embodiment, obstacles and sections of the scanning area obscured by obstacles are determined by evaluating measurement data from a lidar sensor and / or measurement data from at least one camera sensor. This approach provides another option for determining obstacles. Preferably, sensor data fusion with measurement data from another sensor (e.g., a camera sensor or a lidar sensor) can be performed. Optical sensors can, in particular, detect the appearance and size of obstacles.

[0025] According to another embodiment, the masked section of the scanning area is determined as an area Or a volume. Once an obstructing object or obstacle is detected, this information can be used to determine the obscured section of the scanning area. The obscured section can also be determined from a straight-line radar beam that strikes a static or dynamic obstacle. The space or area behind the obstacle forms the obscured section.

[0026] According to another embodiment, a simplified object prediction is created for obscured sections, wherein the simplified object prediction assumes that a potentially dynamic object can cross the vehicle's lane at any time and in any direction. This allows possible reflections or locations in obscured sections of the scanning area to be ignored for collision calculations. Such locations occur as soon as obstructions no longer block the entire radar beam. The radar beam can, for example, radiate through the windows of a parked vehicle or through partially obstructing vegetation, and can be used to obtain information. A simplified, conservative object prediction is assumed. This prediction includes the possibility that an object can escape from the obscured area at any time, in any direction and at any speed, enter the vehicle's field of view, and cross the vehicle's lane. This makes it technically particularly simple to create the object prediction.

[0027] According to another embodiment, the object prediction for the obscured segment is specified using data received from the environment model. To refine the object prediction, it is possible to take into account information about the vehicle environment and create a more accurate object prediction. This data can be obtained, for example, from the environment model and include all aspects that allow for the specification of the existing prediction.

[0028] According to another embodiment, data on intersections, junctions, exits, crosswalks, speed limits, number of lanes, and / or turning lanes are received from an environment model and / or a digital map, wherein the received data is used to specify an object prediction for the obscured section. If the intersection is located behind an obstacle, for example, it can be assumed that the emerging dynamic object moves along its lane at a known maximum speed. A similar assumption can be made when a crosswalk is located in the obscured section.

[0029] According to another embodiment, an object prediction for the obscured section is determined using measurement data acquired by at least one sensor in the obscured section. This allows reflection points in the obscured area to be evaluated and used for an object hypothesis or prediction. It can be assumed that the detected reflection points represent actual objects behind the obstruction, which can quickly escape from the obscured section of the scanning area.

[0030] Based on the corresponding object predictions and the vehicle's motion, a collision check can be performed. The probability of a potential collision depends on the specific prediction. A conservative prediction results in a higher collision probability. A more specific prediction can reduce the number of potential objects in the obscured section and limit their potential motion for collision checking. Therefore, the prediction of obscured objects directly influences the vehicle's behavior while navigating obstacles.

[0031] According to another embodiment, an object prediction is generated based on dynamic objects ascertained from the measurement data or based on the assumption that dynamic objects are present in a hidden section of the scanning area. Dynamic objects can be located in the hidden area and uniquely identified, with their predicted continued travel determined by the object prediction. Such dynamic objects can be ascertained, for example, by receiving data via a vehicle-to-object (Car-2-x) communication link, receiving map data, evaluating sensor measurement data, or similar methods. If no information about the hidden section is available, an object prediction can be generated for possible dynamic objects in the hidden section.

[0032] The method can be used to define the vehicle behavior in the shaded section and, for example, to adapt the vehicle's speed. Alternatively or additionally, emergency braking can be initiated more quickly or in a more controlled manner to avoid potential collisions with dynamic or static objects. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In the following, preferred embodiments of the present invention are explained in more detail based on a greatly simplified schematic illustration. Here, the following is shown:

[0034] Figure 1 A schematic flow chart for explaining a method according to one embodiment is shown;

[0035] Figure 2 A traffic situation is shown for illustrating a simplified object prediction;

[0036] Figure 3 Another traffic situation is shown to illustrate the concrete object forecast. DETAILED DESCRIPTION

[0037] Figure 1 A schematic flow chart is shown for illustrating a method 1 according to an embodiment. Figure 2 and Figure 3 The driving behavior of the vehicle 2 shown in FIG. 1 and can be implemented by the control device 4 . In the exemplary embodiment shown, the control device 4 is designed as a vehicle-side control device 4 .

[0038] Method 1 makes it possible to predict a dynamic object 6 that has escaped from a hidden section 8 or area of ​​a scanning area A of at least one radar sensor 10 and / or lidar sensor 12 of a vehicle 2, thereby enabling targeted situation-adaptive behavior of the vehicle 2 to be initiated.

[0039] Method 1 can be implemented as a software component or a hardware component. Method 1 can serve as a preliminary stage or input variable for collision prediction 20. Movement information and movement prediction 18 of vehicle 2 can be used for collision prediction 20, wherein the operation of collision prediction 20 is not described in detail for the sake of simplicity.

[0040] In method 1 , measurement data collected by at least one radar sensor 1 and / or lidar sensor 12 are received 22 .

[0041] By evaluating the received measurement data, at least one obstacle 14 within scanning area A of radar sensor 10 and / or lidar sensor 12 is ascertained 24 .

[0042] In a further step 26 , a section 8 of scanning region A that is obscured by obstacle 14 is ascertained based on the ascertained obstacle 14 and the installation position of radar sensor 10 and / or lidar sensor 12 .

[0043] Subsequently, based on the determined obscured section 8, an object prediction 28 is generated for the obscured section 8 of the scanning area A. The object prediction 28 can be designed as a simplified object prediction 28. In the simplified object prediction 28, a potential dynamic object 6 can cross the lane of the vehicle 2 at any time and in any direction. This object prediction 28 can preferably be performed when no further information, such as historical trajectory data or the like, is available. The scenario is Figure 2 In this case, the possible trajectories and positions of potential dynamic objects 6 are shown by arrows based on object predictions 28 .

[0044] According to a configuration of method 1, information about the environment 30 of the obscured section 8 can be received. For example, data of an environment model, data of measurement data, data via a communication connection 16 or the like can be received. This additional information of the environment 30 can then be used to specify a simplified object prediction 28. Figure 3 A traffic situation is shown in FIG. 3 , in which such a specified object prediction 32 is used.

[0045] Control device 4 can determine from the map data an intersection in which obscured section 8 is located. Based on curve lane F2, potential or hypothetical dynamic object 6 can only travel through the intersection in three directions, which are taken into account by specified object prediction 32. The maximum possible speed of potential dynamic object 6 can also be limited to the locally customary speed.

[0046] Based on the object predictions 28 , 32 , the driving behavior of the vehicle 2 can be adapted 34 depending on the situation in order to minimize the risk of collision with a potentially dynamic vehicle 6 that crosses the lane F of the vehicle 2 .

[0047] In this case, the driving behavior of the vehicle 2 can be adapted 34 in parallel with the collision prediction 20 . The corresponding data or results of the object predictions 28 , 32 can be provided to the collision prediction 20 .

Claims

1. A method (1) for adapting the driving behavior of a vehicle (2) by means of a control device (4), wherein: - receiving measurement data collected by at least one radar sensor (10) and / or lidar sensor (12), and determining at least one obstacle (14) within a scanning area (A) of the radar sensor (10) and / or lidar sensor (12) by evaluating the received measurement data, - based on the determined obstacle (14) and the installation position of the radar sensor (10) and / or lidar sensor (12), determining a section (8) of the scanning area (A) that is obscured by the obstacle (14), - creating an object prediction for the obscured section (8) of the scanning area (A): a dynamic object (6) can potentially cross the lane (F) of the vehicle (2) from the obscured section (8) of the scanning area (A), - based on the object prediction, adapting the driving behavior of the vehicle (2) according to the situation, Data of a digital radar map are received, wherein when evaluating measurement data collected by the radar sensor (10), a comparison of the collected measurement data with the data of the digital radar map is performed in order to ascertain the obstacle (14) and the section (8) of the scanning area (A) that is obscured by the obstacle.

2. The method according to claim 1, wherein The method (1) is carried out as part of a collision prediction (20), or the determined object prediction and the situation-adapted driving behavior of the vehicle (2) are used as input variables for the collision prediction (20).

3. The method according to claim 1 or 2, wherein: The obstacle (14) and the section (8) of the scanning area (A) that is shaded by the obstacle (14) are ascertained by evaluating measurement data of the lidar sensor (12) and / or measurement data of at least one camera sensor.

4. The method according to claim 1 or 2, wherein: The shaded section (8) of the scanning region (A) is determined as an area or a volume.

5. The method according to claim 1 or 2, wherein: A simplified object prediction is created for the obscured section (8), wherein in the simplified object prediction, potential dynamic objects (6) can cross the lane (F) of the vehicle (2) at any time and in any direction.

6. The method according to claim 1 or 2, wherein: Object predictions for occluded regions are specified by the received data of the environment model.

7. The method according to claim 6, wherein: The data of the environment (30) are received as data of intersections, junctions, exits, crosswalks, speed limits, number of lanes and / or turning lanes from an environment model and / or a digital map, wherein the received data specify an object prediction for the obscured section.

8. The method according to claim 6, wherein: The object prediction for the shaded section (8) is specified by measurement data ascertained in the shaded section by at least one sensor (10, 12).

9. The method according to any one of claims 1, 2, 7 and 8, wherein The object prediction is generated based on a dynamic object (6) ascertained from the measurement data or based on an assumed assumption that a dynamic object (6) is present in a shaded section (8) of the scanning region (A).

10. A control device (4), wherein: The control device (4) is configured to carry out the method according to any one of claims 1 to 9.

11. A computer program product comprising a computer program, the computer program comprising instructions which, when executed by a computer or a control device (4), cause the computer or the control device to carry out the method according to any one of claims 1 to 9.

12. A machine-readable storage medium having a computer program stored thereon, the computer program comprising instructions which, when executed by a computer or a control device (4), cause the computer or the control device to carry out the method according to any one of claims 1 to 9.

Images