阅读说明：本技术 用于为自动化机动车选择操作选项的驾驶系统和方法 (Driving system and method for selecting operating options for an automated motor vehicle ) 是由 C·胡布曼 N·奎奇利希 J·舒尔茨 D·阿尔特霍夫 于 2019-10-25 设计创作，主要内容包括：本发明涉及一种用于机动车(300)的自动驾驶的驾驶系统,其中该驾驶系统被设置为：确定机动车(300)的至少一个传感器的视野(FOV),必要时针对机动车周围环境中的、不在至少一个传感器的视野(FOV)中的至少一个区域(OC1、OC2),模拟在机动车(300)的周围环境中的至少一个区域中的虚拟道路使用者(401),根据机动车(300)与至少一个虚拟道路使用者(401)的可能的虚拟碰撞进行对操作选项(130、131)的评估,据此选择操作选项(130、131)中的一个操作选项,并且执行所选择的操作选项(130、131)。(The invention relates to a driving system for autonomous driving of a motor vehicle (300), wherein the driving system is arranged to: determining a field of view (FOV) of at least one sensor of the motor vehicle (300), optionally for at least one region (OC1, OC2) in the surroundings of the motor vehicle which is not in the field of view (FOV) of the at least one sensor, simulating a virtual road user (401) in the at least one region in the surroundings of the motor vehicle (300), evaluating the operating options (130, 131) as a function of a possible virtual collision of the motor vehicle (300) with the at least one virtual road user (401), selecting one of the operating options (130, 131) as a function thereof, and executing the selected operating option (130, 131).)

1. A driving system for autonomous driving of a motor vehicle (300), wherein the driving system is arranged to:

-detecting (200) an actual traffic situation (100) in the surroundings of the motor vehicle (300),

-determining an actual field of view (FOV) of at least one sensor of the motor vehicle (300) in the actual traffic situation (100),

-simulating, for at least one region (OC1-OC2) in the surroundings of the motor vehicle (300) that is not in the actual field of view (FOV) of the at least one sensor, a virtual road user (401) according to a probability of occurrence of the virtual road user (401) in the at least one region in the surroundings of the motor vehicle (300),

-considering (210) for a first time step at least two alternative operating options (130, 131) relating to longitudinal and/or lateral guidance of the own vehicle (300) based on the actual traffic situation (100),

-determining (220) for the operation options (130, 131), respectively, at least one possible future traffic situation (101- "105") for the first time step as a result of the respective operation option (130, 131),

-determining the spatial position of at least one virtual road user (401) in each possible future traffic situation (101- > 105) in the following manner: the at least one virtual road user (401) moves starting from the spatial position of the at least one virtual road user (401) in the actual traffic situation (100),

-evaluating (250) the possible future traffic situation (101-105) in each case on the basis of a possible virtual collision of the own vehicle (300) with the at least one virtual road user (401),

-selecting (260) one of the operation options (130, 131) for the first time step depending on the evaluation, and

-executing (270) the selected operation option (130, 131).

2. The driving system according to claim 1, wherein the probability of occurrence of the virtual road user (401) depends on the spatial size of an area (OC1-OC2) in the surroundings of the motor vehicle (300) that is not in the actual field of view (FOV) of the at least one sensor.

3. The driving system according to any of the preceding claims, wherein the driving system is arranged to:

-determining a virtual field of view of at least one sensor of the motor vehicle (300) in the possible future traffic conditions (101-

-simulating, for at least one region of the surroundings of the motor vehicle (300) that is not in the virtual field of view of the at least one sensor, a virtual road user (401) according to the probability of occurrence of the virtual road user (401) in the at least one region of the surroundings of the motor vehicle (300).

4. The driving system according to any of the preceding claims, wherein the driving system is arranged to simulate the virtual road user (401) in the following way: the front edge of the virtual road user (401) is located directly at the edge of a region (OC1-OC2) in the surroundings of the motor vehicle (300) that is not in the actual field of view (FOV) of the at least one sensor in the driving direction of the virtual road user (401).

5. The driving system according to any of the preceding claims, wherein the driving system is arranged to simulate the virtual road user (401) in the following way: the length of the virtual road user (401) in the direction of travel along the longitudinal axis is assumed to be infinitely long.

6. The driving system according to any of the preceding claims, wherein the driving system is arranged to: the spatial position of the at least one virtual road user (401) in the possible future traffic situation (101-.

7. The driving system according to claim 6, wherein the constant speed of the virtual road user (401) is higher than the highest speed actually allowed in the respective possible future traffic situation (101-105).

8. The driving system according to any of the preceding claims, wherein the driving system is arranged to:

-determining (230), for at least one of the possible future traffic conditions (101- & 105) for the first time step, one or more operating options (132, 133) related to the longitudinal and/or lateral guidance of the own vehicle (300) based on the respective possible future traffic condition (104) for at least one second time step,

-determining (240), for each of the operational options (132, 133) for the second time step, at least one possible future traffic condition (106) 109 for the second time step as a result of the respective operational option (132, 133) for the second time step,

-determining (245) the spatial position of the at least one virtual road user (401) in each possible future traffic situation (101- > 105) for the second time step in the following manner: the at least one virtual road user (401) is moved starting from the spatial position of the at least one virtual road user (401) in the respective possible future traffic situation (101- "105) for the first time step,

-evaluating (250) each of the possible future traffic conditions (106 and 109) for the second time step as a function of the possible virtual collisions of the own vehicle (300) with the at least one virtual road user (401), and

-selecting (260) the operation option (130, 131) for the first time step based on the evaluation of the possible future traffic conditions (101-.

9. The driving system of claim 8, wherein the driving system is configured to:

-determining a virtual field of view of at least one sensor of the motor vehicle (300) in the possible future traffic conditions (106-,

-simulating a virtual road user (401) in at least one region in the surroundings of the motor vehicle (300) that is not in the virtual field of view of the at least one sensor according to a probability of occurrence for the at least one region in the surroundings of the motor vehicle.

10. The driving system according to any one of the preceding claims,

-the operating option (130) associated with the longitudinal guidance of the own vehicle (300) is to reduce the speed of the own vehicle (300).

11. A method for automatic driving of a motor vehicle, wherein the method comprises the steps of:

-detecting (200) an actual traffic situation (100) in the surroundings of the motor vehicle (300),

-determining (203) an actual field of view (FOV) of at least one sensor of the motor vehicle (300) in the actual traffic situation (100),

-simulating the virtual road user (401) according to the probability of occurrence of the virtual road user (401) in the at least one region (OC1-OC2) in the surroundings of the motor vehicle (300) for at least one region (OC1-OC2) in the surroundings of the motor vehicle (300) that is not in the actual field of view (FOV) of the at least one sensor,

-considering (210) for a first time step at least two alternative operating options (130, 131) relating to longitudinal and/or lateral guidance of the own vehicle (300) based on the actual traffic situation (100),

-determining (220) for the operation options (130, 131), respectively, at least one possible future traffic situation (101- "105") for the first time step as a result of the respective operation option (130, 131),

-determining the spatial position of at least one virtual road user (401) in each possible future traffic situation (101- > 105) in the following manner: the at least one virtual road user (401) moves starting from the spatial position of the at least one virtual road user (401) in the actual traffic situation (100),

-evaluating (250) the possible future traffic situation (101-105) in each case on the basis of a possible virtual collision of the own vehicle (300) with the at least one virtual road user (401),

-selecting (260) one of the operation options (130, 131) for the first time step depending on the evaluation, and

-executing (270) the selected operation option (130, 131).

Technical Field

The invention relates to a driving system and a method for automatic driving of a motor vehicle.

Background

In the context of this document, the term "autonomous driving" may be understood as driving with automatic longitudinal or lateral guidance or autonomous driving with automatic longitudinal and lateral guidance. The term "automatic driving" includes automatic driving with any degree of automation. Exemplary degrees of automation are assisted driving, partially autonomous driving, highly autonomous driving, or fully autonomous driving. The degree of automation is defined by the federal highway institute (BASt) (see the BASt publication "Forschung kompakt", version 11/2012). In assisted driving, the driver continuously performs longitudinal or transverse guidance, while the system takes over the respective other functions within certain limits. In partial automatic driving (TAF), the system takes over longitudinal and transverse guidance for a certain period of time and/or in certain situations, wherein the driver must continuously monitor the system as in assisted driving. In highly automated driving (HAF), the system takes over longitudinal and lateral guidance for a period of time without the driver continuously monitoring the system; the driver must however be able to take over the vehicle control within a certain time. In fully automated driving (VAF), the system can manage driving automatically in all cases for a specific application; the driver is no longer required for this application. The four degrees of automation mentioned above correspond to SAE levels 1 to 4 of the SAE J3016 standard (SAE-automobile engineering society) according to the definition of BASt. For example, high automatic steering (HAF) according to BASt corresponds to level 3 of the SAE J3016 standard. In addition, SAE level 5 is also specified in SAE J3016 as the highest degree of automation, which is not included in the definition of BASt. SAE level 5 corresponds to unmanned driving, where the system can automatically handle all cases like a human driver during the entire trip; the driver is generally no longer required.

Methods for automatic driving of motor vehicles are known. These methods are based on the fact that: the motor vehicle detects its surroundings by means of sensors and controls the vehicle on the basis of the detected sensor data.

However, areas in the surroundings of the motor vehicle may not be visible to the sensor, for example because these areas are obscured by obstacles.

These shadowed areas are problematic for automatic vehicle control because it is not possible to determine unambiguously whether there are other road users that may be relevant for vehicle control.

Disclosure of Invention

The object of the present invention is to provide an improved solution for selecting operating options for an automated motor vehicle, in the surroundings of which there are areas in which sensors of the motor vehicle are not visible.

This object is achieved by the features of the independent claims. Advantageous embodiments are described in the dependent claims. It is noted that additional features of the claims depending on the independent claims may form an invention of their own and independent of a combination of all features of the independent claims, without the features of the independent claims, or in combination with only a subset of the features of the independent claims, as subject matter of independent claims, divisional applications or subsequent applications. This applies in the same way to the technical teaching described in the description which may form the invention independently of the features of the independent claims.

A first aspect of the invention relates to a driving system for automatic driving of a motor vehicle.

The driving system is arranged to detect actual traffic conditions in the surroundings of the motor vehicle.

In particular, the actual traffic situation includes at least one other road user in addition to the own motor vehicle.

The actual traffic situation may in particular comprise spatial information about the own motor vehicle, the road users and/or the traffic infrastructure. The spatial information can be, for example, the position and/or orientation of the vehicle itself or of a road user.

Alternatively or additionally, the actual traffic situation may also comprise, in particular, dynamic information about the state of the road users or the traffic infrastructure. This can be, for example, the activation state of a turn indicator of a road user or the current state of a traffic light as traffic infrastructure.

The detection of the actual traffic situation can take place in particular by means of sensors of the motor vehicle, for example at least by means of a camera, a laser scanner and/or an ultrasonic sensor.

The road user may in particular be a motor vehicle, a cyclist, a pedestrian or any other road user.

The driving system is further arranged to determine an actual field of view of at least one sensor of the motor vehicle in actual traffic conditions.

The field of view of the at least one sensor is essentially determined by the type of construction of the sensor and is limited to objects in the surroundings of the motor vehicle. For example, static or stationary objects in the surroundings of a motor vehicle, such as surrounding buildings or traffic infrastructure beside a roadway, may limit the field of view.

Alternatively or additionally, dynamic objects, such as other road users, may limit the field of view.

At least one region in the surroundings of the motor vehicle, which is not in the actual field of view of the at least one sensor, for example because it is obscured by static or dynamic objects in the surroundings of the motor vehicle, is provided with: for the at least one region in the surroundings of the motor vehicle, virtual road users are simulated as a function of their probability of occurrence in the at least one region in the surroundings of the motor vehicle.

Thus, virtual road users are assumed with a certain probability in at least one area.

Furthermore, the driving system is further configured to: at least two alternative operating options relating to the longitudinal and/or lateral guidance of the own vehicle are considered for the first time step on the basis of the actual traffic situation.

The operating option can in particular be an operating option for changing the vehicle speed and for example an operating option for increasing or decreasing the vehicle speed.

Alternatively or additionally, the operating option may also be an operating option that does not change the speed of the motor vehicle.

Alternatively or additionally, the operating option may be an operating option that changes the lateral orientation of the motor vehicle in the lane.

The operating options describe in particular possible options that can be implemented by the motor vehicle in the actual traffic situation, which influence the longitudinal guidance of the motor vehicle. For example, the operating options of a vehicle are limited by the driving power and physical regulations of the vehicle.

In this case, the operating option can be selected in particular from a predefined set of operating options and parameterized, for example. Alternatively or additionally, the operating options can also be freely determined without prior specification.

In particular, the number of operation options is limited to countable. For example, the driving system is configured to consider less than 10, less than 25, or less than 50 operating options. For each of these operational options, a likely future traffic condition for the first time step is determined as a result of the respective operational option.

The possible future traffic situation may in particular be a hypothetical traffic situation which has not actually occurred at the current point in time. This may in particular comprise spatial information about the own motor vehicle, road users and/or traffic infrastructure, similar to the actual traffic situation.

Alternatively or additionally, the possible future traffic conditions may also comprise dynamic information about the state of the road users or the traffic infrastructure. The possible future traffic situation for the first time step may be derived from the actual traffic situation, in particular by using a stochastic method. For this purpose, the spatial position of the selected road user in the possible future traffic situation for the first time step can be determined, for example, by means of a random vector on the basis of the spatial position of the selected road user in the actual traffic situation.

In the case of possible future traffic situations for the first time step, the spatial position of the own vehicle in particular can be determined as follows: based on the spatial position of the own vehicle in the actual traffic situation, it is assumed that the operating options related to the longitudinal guidance of the own vehicle are executed.

Furthermore, the driving system is further configured to: the spatial position of at least one virtual road user in each possible future traffic situation is determined as follows: the at least one virtual road user moves starting from the spatial position of the at least one virtual road user in the actual traffic situation.

Each of these possible future traffic situations is evaluated as a function of a possible virtual collision of the own vehicle with at least one virtual road user, wherein the evaluation in particular comprises an evaluation of a potential hazard and/or an evaluation of the driving comfort of the own vehicle.

The operating option of the operating options relevant to the longitudinal guidance of the motor vehicle is selected for the first time step. In this case, in particular, an operating option can be selected which corresponds to a traffic situation which indicates the lowest possible risk potential and/or the highest possible driving comfort of the own vehicle.

The selected operating option is then executed in connection with the longitudinal guidance of the own vehicle.

The driving system is particularly configured to: each of the possible future traffic conditions is evaluated by determining at least one cost value associated with the respective possible future traffic condition.

In this case, the cost value can be determined in particular as a function of the spatial position or orientation of the own vehicle and/or of another road user.

Alternatively or additionally, the cost value can also be determined in particular as a function of the spatial relationship of the own motor vehicle, of at least one road user and/or of at least one object in the traffic infrastructure. For example, the cost value may indicate a spatial distance between the own vehicle and a road user. In this case, a high cost value can characterize small spatial distances and a low cost value can characterize large spatial distances, with the other influencing variables being identical. Thus, the cost value may represent, for example, a potential risk for the own vehicle.

The driving system is particularly configured to: the cost value is determined as a function of at least one operating option which leads to a corresponding possible future traffic situation, in particular as a function of the braking process and/or the acceleration process of the respective motor vehicle. For example, high cost values may characterize large numerical variations in vehicle speed. Alternatively, the low cost value may represent a small numerical variation of the speed of the motor vehicle.

Alternatively or additionally, the cost value may also represent, for example, a change in the acceleration of the motor vehicle, wherein both a change in the sign of the acceleration acting on the own motor vehicle and a change in the acceleration value may be taken into account. The cost value can thus be representative of or dependent on the driving comfort of the motor vehicle itself, for example.

The driving system is particularly configured to: the cost value is determined as a function of the respective possible future traffic situation itself, in particular as a function of the collision between the vehicle and the road user in the respective possible future traffic situation. For example, a high cost value may characterize a collision between the own vehicle and a road user. Such a high cost value characterizing a collision may be, for example, at least one order of magnitude greater than all other cost values without a collision.

Alternatively or additionally, the risk of collision, in particular between the own vehicle and the road user, can also be taken into account by the cost value. For example, a high cost value may characterize a collision risk, wherein a cost value characterizing a collision risk may be lower than a cost value characterizing an actual collision.

The driving system is particularly configured to: each of these possible future traffic conditions is determined based on at least one possible movement of the road user.

In particular, the possible movements of the road users can be determined from the degrees of freedom of the road users. For example, if the road user is a motor vehicle, its movement possibilities are limited in particular by its maximum steering angle and its drive power. Alternatively, if the road user is a pedestrian, its movement possibilities are limited by its low movement speed compared to motorised road users.

By determining the likely future traffic situation from at least one possible movement of the road user, in particular only the actual likely future traffic situation that can actually occur within the physical range is taken into account.

Alternatively or additionally, the possible movement of the road user can be determined, in particular, depending on the state of the road user itself. For example, if the road user is a motor vehicle, it can be inferred from the road user's activated steering indicator that the road user is about to turn and in which direction.

In particular, the determination of each of these possible future traffic situations as a function of at least one possible movement of the road user is based on the following idea: in this way, the number of possible future traffic conditions may be reduced compared to the number of all theoretically possible future traffic conditions. Thus, the operating options associated with the longitudinal guidance of the own vehicle can be selected more efficiently according to the invention.

The driving system is particularly configured to: for each of these possible future traffic conditions, a variable is derived that characterizes the probability of occurrence of the respective possible future traffic condition, and an operating option is selected accordingly.

The probability of occurrence of the respective possible future traffic situation can be determined in particular as a function of the probability of the traffic participant changing its spatial position and/or its state.

For example, in the case of traffic conditions at a rotary, the probability of a motor vehicle as a road user leaving the rotary at the next possible exit can be taken into account.

To determine this probability, for example, the spatial orientation of the motor vehicle can be evaluated. If the front of the motor vehicle is oriented in the direction of the next possible exit from the rotary, the probability of the motor vehicle leaving the rotary is higher than if the motor vehicle did not leave the rotary, for example.

Alternatively or additionally, the probability of occurrence of the respective possible future traffic situation may be determined in particular from the probability of an object in the traffic infrastructure changing its state.

For example, for the case of traffic lights it may be assumed that: the longer the current state of the traffic light has existed, the greater the probability of a state change over time.

The probability of occurrence of a traffic situation may in particular also depend on the time. As a result, possible future traffic conditions may become more or less likely over time. For example, the vehicle may move toward an intersection where the vehicle has multiple possibilities of turning. If the vehicle is far away from the intersection, each turn probability is substantially the same. However, once the vehicle enters the intersection and is heading in the direction of one exit of the intersection, it can be assumed that this turn probability is more likely than the other remaining turn probabilities.

The driving system is particularly configured to: variables characterizing the probability of occurrence of the likely future traffic conditions are used as weighting factors for evaluating the respective likely future traffic conditions.

Thus, if the traffic conditions otherwise have the same assessment, then in particular a first possible future traffic condition having a higher probability of occurrence than a second possible future traffic condition may be assessed as being higher than the second possible future traffic condition.

In an advantageous embodiment, the probability of occurrence of a virtual road user is dependent on the spatial size of the region in the surroundings of the motor vehicle which is not in the actual field of view of the at least one sensor.

For example, the probability of occurrence can be determined from the assumed traffic density, such that the probability of occurrence is obtained as a product of the spatial extent of the region in the surroundings of the motor vehicle which is not in the actual field of view of the at least one sensor and the assumed traffic density.

In a further advantageous embodiment, the driving system is provided for: for at least one possible future traffic situation, a virtual field of view of at least one sensor of the motor vehicle in the possible future traffic situation is determined, and for at least one region in the surroundings of the motor vehicle which is not in the virtual field of view of the at least one sensor, virtual road users are simulated as a function of the probability of occurrence of virtual road users in the at least one region in the surroundings of the motor vehicle.

The driving system is particularly configured to: at least one region in the surroundings of the motor vehicle which is not in the virtual field of view of the at least one sensor is associated with a region in the surroundings of the motor vehicle which is not in the virtual field of view of the at least one sensor for a further time step or a region in the surroundings of the motor vehicle which is not in the actual field of view of the at least one sensor.

In particular, if no virtual road user is simulated in the corresponding region of the other time steps, the virtual road user can be simulated in at least one region in the surroundings of the motor vehicle according to the probability of occurrence of the virtual road user.

In an advantageous embodiment, the driving system is arranged to simulate a virtual road user in the following manner: the front edge of the virtual road user is located directly at the edge of a region in the surroundings of the motor vehicle which is not in the actual field of view of the at least one sensor in the direction of travel of the virtual road user.

In an advantageous embodiment, the driving system is arranged to simulate a virtual road user in the following manner: the length of the virtual road user in the direction of travel along the longitudinal axis is assumed to be infinitely long.

This assumption can be implemented, for example, as follows: the length of the virtual road user in the direction of travel along the longitudinal axis is selected to be very large, in particular significantly larger than the length of the real road user. The length of the virtual road user in the direction of travel along the longitudinal axis can thus be greater than 10m, 25m or 100m, for example.

In a further advantageous embodiment, the driving system is provided for: the spatial position of at least one virtual road user in a possible future traffic situation is determined from the constant speed of the virtual road user.

In particular, the constant speed of the virtual road user is higher than the highest speed actually allowed in the respective possible future traffic situation. For example, the constant speed of the virtual road user may be at least 120% or at least 130% of the actually allowed maximum speed.

In the above advantageous embodiments, the driving system is arranged to simulate the virtual road user by assuming an infinitely long length of the virtual road user in the direction of travel along the longitudinal axis, and the constant speed of the virtual road user is higher than the highest speed actually allowed in the respective possible future traffic situation, in particular by the combination of these advantageous embodiments a synergistic effect is obtained which can be taken into account effectively in the "worst case".

An advantage of such a "worst-case" consideration is that all possible states relating to the position and speed of the virtual road user can be covered thereby in terms of the position and speed of the virtual road user relative to regions of the surroundings of the motor vehicle which are not within the field of view of the at least one sensor.

In a further advantageous embodiment, the driving system is provided for: for at least one of these possible future traffic conditions, at least one operating option relating to longitudinal and/or lateral guidance of the own vehicle is determined based on the respective possible future traffic condition for at least a second time step.

In this case, at least one second time step follows the first time step. In particular, the at least one second time step may also be more than just one second time step, wherein the further subsequent time steps correspond to the second time steps. Therefore, only the second time step is explicitly mentioned in this document.

The at least one operating option for the second time step, which is associated with the longitudinal guidance of the own vehicle, can be determined in particular analogously to the operating option for the first time step. The determination of the operating option for the second time step can be distinguished from the determination of the operating option for the first time step in that not the actual traffic situation but a possible future traffic situation forms the initial situation for the determination.

At least one possible future traffic condition for the second time step is determined for each of the operational options for the second time step as a result of the respective operational option for the second time step.

The possible future traffic conditions for the at least one second time step may in particular be determined similarly to the possible future traffic conditions for the first time step.

Thus, by being able to derive the possible future traffic conditions for the second time step from the possible future traffic conditions for the first time step, in particular a chain and/or tree of successively consecutive possible future traffic conditions may be generated.

Determining the spatial position of the at least one virtual road user in each possible future traffic situation for the second time step in the following manner: the at least one virtual road user moves starting from the spatial position of the at least one virtual road user in the respective possible future traffic situation for the first time step, similarly to the movement of the virtual road user between the actual traffic situation and the possible future traffic situation within the first time step.

Each of these possible future traffic conditions for the second time step is evaluated on the basis of a possible virtual collision of the own vehicle with at least one virtual road user, in particular analogously to the evaluation of the possible future traffic conditions for the first time step.

For example, the operational option for the first time step is selected based on an evaluation of likely future traffic conditions for the first time step and based on an evaluation of likely future traffic conditions for the at least one second time step.

In particular, the evaluation of the possible future traffic conditions for the first time step and the evaluation of the possible future traffic conditions for the second time step can be carried out in accordance with a tree structure in which all possible future traffic conditions are associated with one another by means of the operating options.

For example, a sum of the evaluated values for all possible future traffic conditions, which may be caused directly and/or indirectly by the operating options, may be determined for each operating option for the first time step. Thus, the operation option with the highest evaluation for a specific time range can be selected.

Alternatively, for example, for the first time step, an operating option can also be selected which is associated with the longitudinal guidance of the own vehicle and from which the highest rated path starts in the tree structure of possible future traffic situations.

In a further advantageous embodiment, the driving system is provided for: for at least one possible future traffic situation for the second time step, a virtual field of view of the at least one sensor of the motor vehicle in the possible future traffic situation is determined, and for at least one region of the surroundings of the motor vehicle which is not in the virtual field of view of the at least one sensor, a virtual road user in the at least one region of the surroundings of the motor vehicle is simulated as a function of the probability of occurrence.

In a further advantageous embodiment, the operating option relating to the longitudinal guidance of the own vehicle is to reduce the speed of the own vehicle. A reduction in the speed of the motor vehicle can in particular lead to a reduction in the risk of potential hazards and, for example, of collisions with other road users.

A second aspect of the invention relates to a method for automatic driving of a motor vehicle.

One step of the method is detecting an actual traffic situation in the surroundings of the motor vehicle.

Another step of the method is determining an actual field of view of at least one sensor of the motor vehicle in an actual traffic situation.

The method comprises the following steps: for at least one region of the surroundings of the motor vehicle which is not in the actual field of view of the at least one sensor, the virtual road users are simulated as a function of the probability of occurrence of the virtual road users in the at least one region of the surroundings of the motor vehicle.

The method comprises the following steps: at least two alternative operating options relating to the longitudinal and/or lateral guidance of the own vehicle are considered for the first time step on the basis of the actual traffic situation.

The method comprises the following steps: at least one possible future traffic situation for the first time step is determined for each of the operating options as a result of the respective operating option.

Another step of the method is to determine the spatial position of at least one virtual road user in each possible future traffic situation in the following manner: the at least one virtual road user moves starting from the spatial position of the at least one virtual road user in the actual traffic situation.

The method comprises the following steps: these possible future traffic conditions are evaluated in each case on the basis of possible virtual collisions of the own vehicle with at least one virtual road user.

The method comprises the following steps: one of these operation options is selected for the first time step accordingly.

Another step of the method is to perform the selected operation option.

The statements made above with respect to the driving system according to the invention according to the first aspect of the invention also apply in a corresponding manner to the method according to the invention according to the second aspect of the invention. Advantageous embodiments of the method according to the invention according to the second aspect of the invention which are not explicitly specified here and in the claims correspond to the advantageous embodiments described above or in the claims of the driving system according to the invention according to the first aspect of the invention.

Drawings

The invention is explained below with the aid of embodiments with reference to the drawings. Wherein:

figure 1 shows an exemplary tree structure consisting of traffic conditions and operational options,

figure 2 shows an exemplary flow chart of the manner of action of the driving system according to the invention,

figure 3a shows an exemplary traffic situation with another road user,

FIG. 3b shows an exemplary probability curve, an

Fig. 4 shows an exemplary traffic situation with virtual road users.

Detailed Description

Fig. 1 shows an exemplary tree structure consisting of traffic conditions 100 and 109 and operation options 130 and 133.

The actual traffic conditions 100 constitute the root nodes of the tree structure. The actual traffic situation 100 describes, for example, the spatial position of the own vehicle 300 and a plurality of additional road users 310 (for positions see fig. 3a), such as other motor vehicles or pedestrians.

Based on the actual traffic situation 100, the own vehicle 300 can be provided with two different operating options 130, 131 for the first time step with respect to the longitudinal guidance of the own vehicle 300. These operating options may, for example, cause a change in the speed of the motor vehicle 300, for example an acceleration maneuver or a braking maneuver.

If the own vehicle 300 executes the operating option 130, the traffic situation is not yet available only from this, since a change in the spatial position of the additional road user 310 cannot be inferred for the first time step from the operating option 130 alone. For this reason, the exemplary tree structure includes the secondary node 120 and 123.

Based on the auxiliary node 120, different assumptions 140 can be made 142 regarding the change in the spatial position of the additional road user 310 within the first time step, for example by means of a stochastic process or by evaluating the freedom of movement of the additional road user 310. These assumptions can be interpreted, for example, as assumed, non-actual sensor values.

Based on the assumption 140 of the change in the spatial position of the additional road user 310 and the operating options 130 of the own vehicle 300, different possible future traffic situations 101 and 103 for the first time step result.

If the own vehicle 300 executes the operating option 131, assumptions 143, 144 can likewise be made about the change in the spatial position of the additional road users 310. These assumptions 143, 144 may differ from the assumptions 140, 142, for example, although they may also be at least partially the same assumptions.

Based on the assumptions 143, 144 of the change in the spatial position of the additional road user 310 and the operating options 131 of the own vehicle 300, different possible future traffic conditions 104, 105 for the first time step result.

For a second time step following the first time step, the own vehicle 300 can in turn be provided with two operating options 132, 133, for example on the basis of the probable future traffic situation 104.

If the own vehicle 300 executes the operation option 132 in the second time step, the assumptions 145, 146 can again be made about the change in the spatial location of the additional road user 310 within the second time step based on the likely future traffic conditions 104 for the first time step.

Based on the assumptions 145, 146 about the change in the spatial position of the additional road user 310 and the operating options 132 of the own vehicle 300, different possible future traffic conditions 106, 107 for the second time step result. If the own vehicle 300 executes the operating option 133 in the second time step, an assumption 147, 148 can likewise be made about the change in the spatial position of the additional road user 310 within the second time step on the basis of the probable future traffic situation 104 for the first time step.

Based on the assumptions 147, 148 of the change in the spatial position of the additional road user 310 and the operating options 133 of the own vehicle 300, different possible future traffic conditions 108, 109 for the second time step result.

Fig. 2 shows an exemplary flow chart of the mode of action of the driving system according to the invention.

In step 200, the actual traffic situation 100 in the environment of the own vehicle 300 is detected, for example, by means of a sensor system of the own vehicle 300.

In step 203, the actual field of view FOV of the at least one sensor of the motor vehicle 300 in the actual traffic situation 100 is determined.

The field of view of the at least one sensor is thereby essentially determined by the type of construction of the sensor and is limited to objects 400 in the surroundings of the motor vehicle, for example to static or dynamic objects 400 in the surroundings of the motor vehicle.

Subsequently, in step 207, for at least one region OC1-OC2 in the surroundings of the motor vehicle 300 which is not in the actual field of view FOV of the at least one sensor, the virtual road user 401 is simulated according to the probability of occurrence of the virtual road user 401 in the at least one region OC1-OC2 in the surroundings of the motor vehicle 300.

In step 210, at least two alternative operating options 130, 131 relating to the longitudinal guidance of the own vehicle 300 are considered for a first time step on the basis of the actual traffic situation 100. The at least two alternative operating options may for example be selected from a predetermined set of possible operating options.

These operating options may be, for example, increasing 130 the speed of the own vehicle 300 and decreasing 131 the speed of the own vehicle 300.

In step 220, the possible future traffic conditions 101- "105 for the first time step are determined for each of these operation options. The possible future traffic situations 101-105 corresponding to the operating options 130, 131 may in particular describe different spatial positions of the other road users 310, respectively.

The spatial position of the other road users 310 can in particular be determined by assuming possible observable movements of the other road users 310. If the road user 310 is located in a roundabout, for example, the possible observable movement of the road user 310 is to leave 144 the roundabout at the next exit or to remain 143 in the roundabout. The two possible observable motions 143, 144 may, for example, each include one of the possible future traffic conditions 104, 105.

In step 225, the spatial position of at least one virtual road user 401 in each of the possible future traffic situations 101-: the at least one virtual road user 401 moves starting from the spatial position of the at least one virtual road user 401 in the actual traffic situation 100.

In a subsequent step 230, the operating options 132, 133 relating to the longitudinal guidance of the own vehicle 300 are taken into account for a second time step. Here, these operating options 132, 133 are based, for example, on the probable future traffic situation 104 at the first time step.

In step 240, similar to step 220, a determination is made of the likely future traffic conditions 106 and 109 for each of these operational options 132, 133 based on the likely observable motions 145 and 148 of the other road users 310.

In step 245, the spatial position of at least one virtual road user 401 in each of the possible future traffic conditions 108 and 109 for the second time step is determined by: the at least one virtual road user 401 moves starting from the spatial position of the at least one virtual road user 401 in the respective possible future traffic situation 101-105 for the first time step.

In step 250, all possible future traffic conditions 101-.

In particular, characteristics of the possible future traffic conditions 101-109 (e.g., distance of the own vehicle 300 and other road users 310) may be included in the evaluation. Since a large distance of the own vehicle 300 and other road users 310 can be interpreted as a very safe state, the large distance is evaluated relatively positively. In contrast, for example, a collision of the own vehicle 300 with the road user 310 can be evaluated relatively negatively.

If the evaluation is carried out, for example, by means of a cost function, the cost value can be inversely related to the distance between the own vehicle 300 and the other road users 310, for example. Therefore, a large distance results in a low cost value. In this case, a collision of the own vehicle 300 with the road user 310 may result in a very high cost value, for example.

Alternatively or additionally, the operation options 130 and 133 leading to the respective future traffic conditions 101 and 109 may also be included in the evaluation. For this purpose, in particular, the influence of the operating options 130 and 133 on the driving comfort of the own vehicle 300 can be determined. For example, large numerical changes in the speed of the own vehicle 300 are evaluated rather negatively in terms of driving comfort. For example, in such a case, a relatively high cost value may be factored into the evaluation of the likely future traffic conditions 101-109. Alternatively, in the case of small variations in the value of the speed of the own vehicle 300, only relatively low cost values may be included in the evaluation of the possible future traffic conditions 101-109.

Alternatively or additionally, in particular the probability of occurrence of the possible future traffic conditions 101-. The following can thus be considered: the very high probability of possible future traffic conditions 101-.

In step 260, the operation options 130, 131 for the first time step are selected according to step 250. For this purpose, in particular, an evaluation of possible future traffic situations which can be traced back to the operating options 130, 131 can be combined for each operating option 130, 131. For example, for the first operation option 130, the evaluation of the possible future traffic conditions 101 and 103, which can be traced back to the first operation option, can be added. For the second operating option 131, the evaluation of the possible future traffic conditions 104 and 109, which can be traced back to the second operating option, can likewise be added.

If the evaluation is for example a cost value, the operation option 130, 131 with the lowest overall cost value may be selected for the first time step. This may have a significant influence on the selection of the operating options 130, 131 relating to the longitudinal guidance of the own vehicle 300, in particular if a collision of the own vehicle 300 with another road user 310 may occur in the selected possible future traffic situation 107. If a possible future traffic situation 107 is assigned a very high cost value, for example due to a collision, which is at least an order of magnitude greater than the cost values of all other possible future traffic situations 101 and 106, 108, 109, it is very unlikely in this way that an operating option 131 is selected which operating option 131 may lead to the selected possible future traffic situation 107.

In particular, this implicitly leads to: if there is at least one possible future traffic situation 107 in the tree structure in which the own vehicle 300 may collide with another road user 310, the driving system according to the invention selects an operating option 130, 131 which is relatively cautious and/or reduces the speed of the own vehicle 300.

The operating options 130, 131 for reducing the speed of the own vehicle 300, for example, have the following effect again: the time until a collision may actually occur is extended, thereby obtaining more time again for more accurately determining the actual behavior of other road users 310.

In step 270, the operation options 130, 131 selected in step 260 are executed for a first time step. The longitudinal guidance of the vehicle 300 is influenced in a manner predetermined by the selected operating options 130, 131. For example, the speed of the own vehicle 300 may be increased 130 or decreased 131.

As soon as the second time step actually occurs, the driving system according to the invention can be operated again, in particular starting from step 200, during which the actual traffic situation 100 has changed significantly. Alternatively or additionally, however, at least the intermediate results of the individual process steps for the first time step can also be stored in a memory module comprised by the driving system and reused when the driving system is re-operated in the actual second time step.

Fig. 3a shows an exemplary traffic situation for illustrating the driving system according to the invention.

Here, the own vehicle 300 is located in front of the intersection into which the vehicle 310 enters from the left side as another road user.

The own vehicle 300 is, for example, intended to cross an intersection on a lane 301, and it is not possible for the own vehicle 300 to determine unambiguously whether other road users 310 will cross the intersection on a lane 311, or whether other road users 310 will turn right onto a lane 312.

Fig. 3b shows an exemplary time curve of the probability that can be evaluated by the driving system according to the invention. The curve 321 indicates the time-dependent probability of the other road user 310 crossing the intersection straight on the lane 311, and the curve 322 indicates the time-dependent probability of the other road user 310 turning right on the lane 312. During the time plotted on the abscissa, the probability that can be evaluated by the driving system varies in the following way: up to the time t of 9s, the two possible movement paths 311, 312 of the other road users 310 appear to be substantially equal with probability. Starting from the time t of 9s, the probability 321 that the other road users 310 decide to select the route 311 passing through the intersection rises sharply.

The reason for this may be, for example, that the other road users 310 are already very close to the intersection and have not yet activated the turn indicator all the time, or that the behavior of the other road users 310 must be recognizable as the situation progresses. From which it can be deduced that the other road users 310 want to go straight.

The movement path 311 of the road user 310 directly across the intersection may lead to a possible future traffic situation 107 in which the own vehicle 300 is about to collide with the road user 310.

Since such possible future traffic situations 107 have, for example, a very high cost value, the driving system according to the invention will tend to select 260 the operating options 130, 131 which do not lead to such possible future traffic situations 107 with a high probability until the time t-9 s. The operating options 130, 131 tend to be operating options 130, 131 that at least do not increase the speed of the vehicle 300. In particular, the operating options 130, 131 are operating options 130, 131 that reduce the speed of the vehicle 300 to produce a look-ahead behavior until further operating options become apparent.

Fig. 4 shows an exemplary traffic situation with virtual road users.

Here, as the actual traffic situation (100), the own vehicle 300 is located before the entrance of the branch, that is, at the intersection of a road and a road without a break without crossing the continuation of the road.

Here, the driving system for the automatic driving of the motor vehicle 300 according to the present invention is configured to: an actual traffic situation 100 in the surroundings of the motor vehicle 300 is detected 200 and an actual field of view FOV of at least one sensor of the vehicle 300 in the actual traffic situation 100 is determined.

Here, the roadside object 400 limits the field of view FOV of the at least one sensor.

Thus, the driving system is arranged to: for at least one region OC1-OC2 in the surroundings of the motor vehicle 300 which is not in the actual field of view FOV of the at least one sensor, the virtual road user 401 is simulated according to the probability of occurrence of the virtual road user 401 in the at least one region in the surroundings of the motor vehicle 300.

Further, the driving system is configured to: at least two alternative operating options 130, 131 relating to the longitudinal and/or lateral guidance of the own vehicle 300, such as for example a speed hold or a speed reduction, are considered 210 for a first time step on the basis of the actual traffic situation 100.

The driving system determines for these operation options 130, 131 at least one possible future traffic situation 101-105 for the first time step as a result of the respective operation option 130, 131, respectively.

Furthermore, the driving system is arranged to determine the spatial position of at least one virtual road user 401 in each of the possible future traffic conditions 101-105 in the following manner: the at least one virtual road user 401 moves starting from the spatial position of the at least one virtual road user 401 in the actual traffic situation 100.

For example, the virtual road user may move at a constant speed toward the branch entrance.

Thus, the driving system is arranged to: these possible future traffic situations 101-105 are evaluated 250 in each case on the basis of possible virtual collisions of the own vehicle 300 with at least one virtual road user 401.

For example, the possible virtual traffic situation resulting from the maintained speed can be evaluated "worse" than the possible virtual traffic situation resulting from the reduced speed, since the risk of a collision of the motor vehicle 300 with the virtual road user 401 is higher in this case.

The driving system accordingly selects 260 one of these operation options 130, 131 for the first time step and executes 270 the selected operation option 130, 131.

For example, the driving system selects the operational option 130, 131 that brings a possible virtual traffic condition that is assessed as "better".

For example, the speed of the motor vehicle 300 is reduced by the necessary amount by continuously and periodically repeating the process to prevent the motor vehicle 300 from colliding with actual other road users that may be in at least one region OC1-OC2 in the surroundings of the motor vehicle 300 that is not in the actual field of view FOV of the at least one sensor.

However, the speed of the motor vehicle 300 is not reduced more than this necessary, since the region OC1-OC2 is reduced during the forward movement of the motor vehicle 300, whereby the risk of collision of the motor vehicle 300 with the virtual road user 401 will be reduced, whereby operating options that do not reduce the speed of the motor vehicle, for example, are evaluated as "better". This can be achieved in particular by: other criteria, in particular the deviation of the speed of the motor vehicle 300 from the target speed, are taken into account in the evaluation. For example, a deviation of the speed of the vehicle 300 from the target speed has a negative effect on the evaluation, which will result in the vehicle 300 trying to reach the target speed when there is no or only a low risk of collision between the vehicle 300 and the virtual road user 401.