DE102012009555A1 - Method for assisting driver during guiding vehicle in crossing area, involves detecting objects present in surrounding of vehicle and determining crossing situation - Google Patents

Abstract

The method involves detecting objects (O1,O2) present in a surrounding of a vehicle (F) and determining a crossing situation. The free surface areas traversable by the vehicle is detected in the surrounding in a flowing traffic. A probability is determined whether the vehicle is guidable collision free in the free surface areas in response to a predicted trajectory (TF-1 to TF-n) of the vehicle and predicted trajectories (TO1-1 to TO1-m,TO2-1 to TO2-z) of collision relevant objects present in the surrounding. The behavior of the objects is considered during the determination of probability. An independent claim is included for a driver assistance system for executing the driver assisting method.

Description

The invention relates to a method for assisting a driver when driving a vehicle in an intersection area, wherein objects located in an environment of the vehicle are detected and an intersection situation is determined.

The invention further relates to a driver assistance system for carrying out such a method.

Driver assistance systems are generally known from the prior art, which detect objects approaching an intersection. According to the EP 1 269 445 B1 the driver assistance systems trigger a collision warning or brake intervention if a prediction reveals that the system vehicle equipped with the driver assistance system and one of the detected objects will be in the intersection at the same time.

Since at high traffic density there is a high probability that the system vehicle and an object will be in the intersection at the same time, such driver assistance systems largely trigger a warning at high traffic density.

From the DE 10 2011 106 176 A1 a method is known for determining a safety zone in a crossing situation. The method is based on a prediction of future movement trajectories of the own vehicle, ie the system vehicle, and a collision-relevant further vehicle. Motion hypothesis trajectories are calculated for the two vehicles. The motion hypothesis trajectories are expected trajectories for various individual maneuver motions of the vehicles. The calculation is carried out taking into account a driving style of an average driver descriptive control profile, which has been determined empirically and for an average driver limits longitudinal acceleration, lateral acceleration and a steering angle velocity indicates that the driver would not exceed normal driving behavior without extreme maneuvers in an intersection situation. From the set of calculated motion hypothesis trajectories those trajectory pairs are identified in which a predetermined safety distance is maintained between the vehicles. The sought security zone is defined by the set of trajectory pairs thus determined.

The invention has for its object to provide a comparison with the prior art improved method for assisting a driver when driving a vehicle and an improved driver assistance system.

With regard to the method, the object is achieved by the features specified in claim 1 and in terms of the driver assistance system by the features specified in claim 10.

Advantageous embodiments of the invention are the subject of the dependent claims.

In the method for assisting a driver when driving a vehicle in an intersection area, objects located in an environment of the vehicle are detected and an intersection situation is determined.

According to the invention, in the environment in a flowing traffic of the vehicle passable free areas are detected, depending on at least one predicted trajectory of the vehicle and predicted trajectories located in its environment collision relevant objects, a probability is determined whether the vehicle collision free in the free areas flow of traffic is feasible, whereby in determining the probability of a classified as cooperative or uncooperative behavior of the objects is taken into account. If a predefined desired probability is undershot, a driver warning is output and / or an automatic longitudinal and / or lateral control of the vehicle to avoid a collision is carried out.

In this way, a collision warning and / or the automatic control are triggered only if the driver of the vehicle, even taking into account a cooperative behavior of the other road users, will not be able to safely pass the intersection.

Advantageously, the desired likelihood is adapted to the condition of a road surface on which the vehicle is located, whereby, given a slippery road surface, less dynamic driving maneuvers are taken as the basis for trajectory prediction than for a grip road surface. The desired likelihood is modified by adaptation to the condition of the road surface such that it is reduced in a situation with a smooth road surface compared to a situation with a grip road surface. This takes into account the fact that the controllability of the vehicle with a slippery road surface is less than with a grip road surface, and therefore the probability that the intersection can be passed without collision, is less with a slippery road surface than with a grip road surface. The condition of the road surface can be determined in a conventional manner by a friction coefficient estimation and / or by evaluation of the Signals of a rain sensor and / or temperature sensor can be determined.

Advantageously, additionally or alternatively, an adaptation of the desired likelihood to the driving style of the driver can be made, wherein the driving style can be determined automatically based on the dynamics of the previous driver behavior or can be specified by the driver himself via an input as driving style specification. In a sporty driving style, trajectory prediction takes into account more dynamic driving maneuvers than a comfort-oriented or consumption-optimized driving style. The desired likelihood is modified by adapting to the driving style such that it is lower in a sporty driving style than in a comfort-oriented or consumption-optimized driving style. This takes into account the fact that in a sporty driving style the risk is higher that the vehicle is in an uncontrollable state, as in a comfort-oriented or consumption-optimized driving style, and thus that the probability that the intersection can be passed without collision, at a sporty driving style is lower than in a comfort-oriented or consumption-optimized driving style.

In particular, cooperative behavior is only assumed for a collision-relevant vehicle if there are recognizable signs that allow the conclusion that the driver of that vehicle will behave in a cooperative manner.

By means of the method according to the invention, it is possible in a particularly advantageous manner to assist a driver of the vehicle even in a high traffic density when safely crossing the intersection. By passing the intersection, all driving maneuvers are understood that the driver can perform in order to leave the intersection area. These include, in particular, crossing under straight-ahead driving or turning off and arranging in the flowing traffic.

In particular, a collision avoidance or at least a collision reduction in all Einschersituationen in left and right traffic of intersections, in the case of driving through intersections in the shearing and the previous temporary stopping and starting the vehicle is possible.

Embodiments of the invention are explained in more detail below with reference to drawings.

Showing:

1 schematically a crossing situation with several vehicles and

2 schematically a driver assistance system for assisting a driver when driving a vehicle in an intersection area.

In 1 is a crossing situation with a vehicle F and two collision-relevant objects O1, O2 shown. The collision-relevant objects O1, O2 are likewise vehicles and / or moving groups of objects O1, O2.

To assist a driver in driving the vehicle F in the intersection area, this includes a driver assistance system 1 , For the sake of clarity, the driver assistance system 1 with its components detailed in 2 shown.

The driver assistance system 1 includes an environment detection device 2 for detecting an environment of the vehicle F and the objects O1, O2 and / or moving groups of objects O1, O2 located therein.

Furthermore, an evaluation device 3 for evaluation by means of the environment detection device 2 acquired environmental data D provided. The environment detection device 2 includes at least one long-range radar, at least one near-field radar and / or at least one camera. From the environmental data D by means of the evaluation device 3 determined in a flowing traffic of the objects O1, O2 of the vehicle F accessible surface areas.

In addition, by means of the evaluation unit 3 performed a prediction to determine if for. the driver of the vehicle F has the ability to control this by accelerating, braking and steering so that he can run the vehicle F through or in the flowing with the flowing traffic free areas and thus can pass through the junction collision. That is, a probability is determined whether the vehicle F is collision-free in the free surface areas of the flowing Verkehrsführbar. If this possibility does not exist, that is, if a predetermined target probability is exceeded, a collision warning is issued to the driver of the vehicle. For this purpose, the evaluation device 3 with a warning device 4 coupled to the output of an optical, acoustic and / or haptic driver warning. In addition, the evaluation device 3 with a control device 5 for automatic longitudinal and / or transverse control of the vehicle F coupled. By means of this, for example, an autonomous braking intervention is triggered if the possibility collision-free crossing of the intersection does not exist. It is essential that an optionally existing cooperative behavior of the other road users is recognized and taken into account in the prediction.

In the prediction, possible trajectories TF ₁ to TF _{n of} the vehicle F and possible trajectories TO1 ₁ to TO1 _m and TO2 ₁ to TO2 _{z of} the objects O1, O2 are predicted.

The following is an expiry of a possible embodiment of the inventive method for assisting the driver when driving the vehicle F is explained.

In a first method step, an intersection situation is first determined. In this case, by means of the environment detection device 2 the objects O1, O2, which are located within a predefinable look-ahead horizon of, for example, 2 seconds in an area in the direction of travel in front of the vehicle, detected and by means of the evaluation device 3 classified as collision-relevant vehicles.

In a second method step by means of the evaluation device 3 from the environmental data D detects the passable surface areas in the vicinity of the intersection as possible driving corridors. The detection is based on the detection of a position and direction of movement of individual moving objects O1, O2, of standing objects, not shown, such as edge structures, curbs and turfs and groups of moving objects O1, O2. Alternatively or additionally, the driving corridors are also determined by detecting lanes by means of a navigation device and by means of a digital road map.

In a third method step, in order to predict the trajectory of the vehicle F from a steering angle, an actuation of a direction indicator and / or a planned route of the navigation device, a target position of the vehicle F for passing the intersection is predicted.

Furthermore, for the prediction of the trajectory of the vehicle F in a fourth method step for the vehicle F, an actual state variable X _is determined. This is a multidimensional vector variable which comprises, as components, actual values of a longitudinal acceleration a _x , a lateral acceleration a _y and a steering angular velocity dδ / dt according to the following equations: X _is = {x ₁ , x ₂ , x ₃ } (1) With x ₁ = a _x , (2) x ₂ = ay, (3) x ₃ = dδ / dt. (4)

Alternatively to the use of the steering angular velocity dδ / dt, x _{3 is obtained} from the yaw rate according to the following equation: x ₃ = dψ / dt. (5)

For the vehicle continues to be a predetermined number of desired state variables X _{soll, i} specified. This is based on the actual state quantity X _is also a multi-dimensional vectorial variable which as components setpoint values of the longitudinal acceleration a _{x, i} , the lateral acceleration a _{y, i} and the steering angular velocity dδ _i / dt according to the following equations: X _{soll, i} = {x _{1, i} , x _{2, i} , x _{3, i} } (6) With x _{1, i} = a _{x, i} , (7) x _{2, i} = ay, i, (8) x _{3, i} = dδ _i / dt, (9) with: i = 1 to y

As an alternative to using the steering angle speed dδ / dt, x _{3, i} results from the yaw rate dψ _i / dt according to the following equation: x _{3, i} = dψ _i / dt. (10)

The desired state variables X _{soll, i} are values that are within predefined limits.

The specification of the limits is based on a driver profile, which describes the typical driving style of an average driver. The driver profile preferably corresponds to that in the DE 10 2011 106 176 A1 profile as an individual control profile ICP and is determined empirically as described there. Through empirical measurements in different countries different driver profiles can be determined for the different countries, which characterize a driving style typical in the respective country. Thus, for example, for the Asian region adapted to the traffic situation in Asian cities driver profile can be determined and used to determine the desired state variables X _{soll, i} .

These limits are further modified depending on a condition of a road surface, so that when driving on slippery road surface less dynamic driving maneuvers are considered than on rough roads. Likewise, a driving style specification made by the driver himself is taken into account in determining the limits in such a way that dynamic driving maneuvers are permitted in a sporty driving style than in a comfort-oriented or consumption-optimized driving style.

The various values of the desired state variables X _{soll, i} take into account the various possible options for action, such as evading, braking and / or accelerating the driver. In particular, not all physically possible trajectories TF ₁ to TF _{n of} the vehicle F are calculated by defining the limits on the basis of the predefined driver profile, but only the trajectories TF ₁ to TF _n typical for an intersection situation in the respective country. As a result, a computational effort is kept low. By considering country-specific driver profiles, an assistance function of the driver assistance system is adapted to the traffic conditions in the respective country 1 realizable.

_Is the basis of the actual state quantity X is _intended for each of the desired state variables _X, a trajectory TF ₁ to TF _n calculated _I which X _{is intended} for a regulation of the respective target state _{quantity, i} results and along the vehicle F in the desired direction can be led over or into the intersection. This calculation is based on a vehicle model, as for example in equation (1) of DE 10 2011 106 176 A1 is described as an extended single-track model, and for the given forecast horizon. Trajectories TF ₁ to TF _{n of} the vehicle F and target state variables X _{soll, i} , which would lead the vehicle F out of the range of drivable areas, are rejected as invalid. Thus, a set of trajectories having a plurality of trajectories TF ₁ to TF _n for the vehicle F is predicted, taking into account the driver's various options for action.

The group of trajectories of the multiplicity of trajectories TF ₁ to TF _n corresponds in particular to those in FIG DE 10 2011 106 176 A1 trajectories called motion hypothesis trajectories.

In a fifth method step, the behavior of the collision-relevant objects O1, O2 are classified. From the means of the environment detection device 2 , in particular by means of at least one camera detected environmental data D is in one by means of the evaluation device 3 Image processing realized whether the driver of a detected as a collision-relevant vehicle object O1, O2 has indicated that he will behave cooperatively. Indications indicating cooperative behavior include, for example, eye contact, certain gestures, such as hand signals and nods, and lights generated by vehicle lighting. Indices of the cooperative behavior of the objects O1, O2 can also be derived from changes in their state of motion. This includes in particular a noticeable slowing down of the objects O1, O2 and evasive maneuvers of the same. These quantities are detected by detecting a reduction in the speed of the objects O1, O2, by detecting an evasion to adjacent lanes and / or by detecting the generation of a free area by the objects O1, O2, around the vehicle F into the flowing traffic or the Driving through it to determine possible determined.

Furthermore, it is checked whether the objects O1, O2 are actually capable of releasing a surface area, that is to say a gap, for the vehicle F by means of a deceleration and / or an evasion and thus a collapsing maneuver for the vehicle F by cooperative To allow behavior. If such indications exist, the behavior of the respective object O1, O2 is classified as cooperative.

Otherwise, or if there are clear indications of non-cooperative behavior, such as a stubborn look forward and / or no change in state of motion to create a free area for a lurching or passing maneuver, the behavior is classified as non-cooperative.

In the illustrated embodiment, the object O1 shows an uncooperative behavior. By contrast, the second object O2 shows a cooperative behavior and can enable the vehicle F to enter the flowing traffic by changing the lane and / or by slowing down its speed.

In a sixth method step, the prediction of the trajectories TO1 ₁ to TO1 _m and TO2 ₁ to TO2 _{z is carried out} for the collision-relevant objects O1, O2.

From the means of the environment detection device 2 detected environmental data D are using the evaluation device 3 in addition to the detection of the collision-relevant objects O1, O2 also determines their positions and states of motion. The environment detection device 2 In one embodiment, this additionally includes a vehicle-to-vehicle communication system and / or an infrastructure-to-vehicle communication system. With such communication systems, it is possible to calculate the position and movement data, in particular, to detect the speed, acceleration, orientation and yaw rate of the objects O1, O2 with high accuracy.

On the basis of the determined position data and movement state data of the collision-relevant objects O1, O2, their trajectories TO1 ₁ to TO1 _m and TO2 ₁ to TO2 _{z of} the objects O1, O2 are determined for the same look-ahead horizon as for the vehicle F. The calculation is carried out either under a first assumption that the respective collision-relevant object O1, O2 is maintained unchanged in uncooperative behavior its movement state or under a second assumption that the respective collision-relevant object O1, O2 in cooperative behavior its speed within predetermined limits will modify and / or will dodge within certain limits to allow the vehicle F to pass the intersection.

The calculation according to the first assumption then takes place when the behavior of the driver of the object O1, O2 designed as a vehicle has been classified as non-cooperative, whereas the calculation according to the second assumption takes place when the behavior of the driver concerned is classified as cooperative has been.

In the calculation according to the second assumption, several potentially possible cooperative reactions, such as light, medium, and strong braking, and / or easy, mediocre, and strong steering are given, and for each of these reactions a separate trajectory TO1 ₁ to TO1 _m and TO2 ₁ to TO2 _z calculated. Thus, for the "cooperative" objects O1, O2, in the illustrated embodiment for the object O2, a trajectory set of a plurality of trajectories TO2 ₁ to TO2 _{z are} determined, which represent the various possibilities of the cooperative behavior. For the "uncooperative" object O1, on the other hand, only a single trajectory of several possible trajectories TO1 ₁ to TO1 _{m is} determined.

In a seventh method step, a situation analysis is performed by analyzing the trajectories TF ₁ to TF _n predicated for the vehicle F and the trajectories TO1 ₁ to TO1 _m and TO2 ₁ to TO2 _z predicted for the collision-relevant objects O1, O2 and the positions which the vehicle F and the collision-relevant objects O1, O2 at different times on their trajectories TF ₁ to TF _n and trajectories TO1 ₁ to TO1 _m , TO2 ₁ to TO2 _{z will} take. In the analysis, it is checked whether the trajectory of the vehicle F contains a trajectory TF ₁ to TF _n , according to which the vehicle F can pass the intersection without collision.

If no such trajectory TF ₁ to TF _{n is} present, in a eighth method step, the system response of the driver assistance system 1 the visual, acoustic and / or haptic collision warning is issued to the driver of the vehicle F and / or the autonomous braking intervention is triggered to prevent the entry into the collision-prone crossing area.

In an advantageous embodiment, if there is at least one collision-free mobile trajectory TF ₁ to TF _n for the vehicle F, an indication is sent to the driver to inform him of the possibility of collision-free passing of the intersection and / or it will be found collision mobile trajectory TF TF ₁ to _n or one of the retrieved collision mobile trajectories TF TF ₁ to _n as a target trajectory for an autonomous control longitudinal and transverse selected. In the context of this autonomous longitudinal and transverse control, the vehicle F by means of the control device 5 according to the desired trajectory autonomously guided by appropriate autonomous interventions in the longitudinal and lateral dynamics of the vehicle F in or through the intersection.

LIST OF REFERENCE NUMBERS

1

Driver assistance system

2

Ambient detection device

3

evaluation

4

warning device

5

control device

D

Environmental data

F

vehicle

O1

object

O2

object

TF ₁ to TF _n

trajectory

TO1 ₁ to TO1 _m

trajectory

TO2 ₁ to TO2 _z

trajectory

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• EP 1269445 B1 [0003]
• DE 102011106176 A1 [0005, 0036, 0039, 0040]

Claims (10)

Method for assisting a driver in driving a vehicle (F) in an intersection area, wherein objects (O1, O2) located in an environment of the vehicle (F) are detected and an intersection situation is determined, characterized in that - in the surroundings in one floating traffic areas can be detected by the vehicle (F), in which at least one predicted trajectory (TF ₁ to TF _n ) of the vehicle (F) and predicted trajectories (TO1 ₁ to TO1 _m , TO2 ₁ to TO2 _z ) a probability is determined by collision-relevant objects (O1, O2) in the vicinity thereof, whether the vehicle (F) can be guided into the free surface areas of the flowing traffic without collision, - wherein, in determining the probability, a behavior of the objects classified as cooperative or uncooperative (O1, O2) is taken into account and - if it falls below a predetermined target probability A driver warning is issued and / or an automatic longitudinal and / or lateral control of the vehicle to avoid a collision is performed. A method according to claim 1, characterized in that the predetermined desired likelihood is adapted to a determined state of a road surface and / or to a predetermined driving style. A method according to claim 1 or 2, characterized in that as collision-relevant objects (O1, O2) objects (O1, O2) are detected, which are located within a predetermined lookahead horizon in an area in the direction of travel in front of the vehicle (F). Method according to one of the preceding claims, characterized in that the passable free surface areas in an environment of the intersection as possible, between stationary objects, moving objects (O1, O2) and / or moving groups of objects (O1, O2) located driving corridors be detected. Method according to one of the preceding claims, characterized in that in the prediction of the trajectory (TF ₁ to TF _n ) of the vehicle (F) from a steering angle, an operation of a direction indicator and / or a planned route of a navigation device, a target position of the vehicle (F ) is predicted after passing the intersection, and / or that in the prediction of the trajectory (TF ₁ to TF _n ) of the vehicle (F) multidimensional vector actual state variables are determined, wherein the actual state variables from actual values of a vehicle longitudinal acceleration , a vehicle lateral acceleration, a steering angle speed and / or a Fahrzeuggierrate are formed, starting from the actual state variables for multidimensional predetermined vectorial desired state variables based on a vehicle model in each case a trajectory (TF ₁ to TF _n ) of the vehicle (F) is determined. A method according to claim 5, characterized in that the vehicle (F) from a free surface area outgoing trajectories (TF ₁ to TF _n ) and target state variables are rejected as invalid. Method according to one of the preceding claims, characterized in that a classification of the behavior of the objects (O1, O2) in dependence on light signals of the objects (O1, O2), eye contact of the objects (O1, O2), gestures of the objects (O1, O2), a possible change of a movement state of the objects (O1, O2) and / or an actual change of a movement state of the objects (O1, O2) is performed. Method according to one of the preceding claims, characterized in that the predicted trajectories (TO1 ₁ to TO1 _m , TO2 ₁ to TO2 _z ) of the objects (O1, O2) within a predetermined look-ahead horizon from position data and / or movement data of the objects (O1, O2 ) under a first assumption in uncooperative behavior in which the respective object (O1, O2) maintains its state of motion unchanged, and under a second assumption in cooperative behavior in which the respective object (O1, O2) changes its state of motion , Method according to one of the preceding claims, characterized in that in a situation analysis based on an analysis of the predicted trajectories (TF ₁ to TF _n ) of the vehicle (F) and the predicted trajectories (TO1 ₁ to TO1 _m , TO2 ₁ to TO2 _z ) of Objects (O1, O2) as well as positions occupied by the vehicle (F) and the objects (O1, O2) at different times on their trajectory (TF ₁ to TF _n , TO1 ₁ to TO1 _m , TO2 ₁ to TO2 _z ) , Trajectories (TF ₁ to TF _n ) are determined on which a collision-free passage of the intersection for the vehicle (F) is executable. Driver assistance system ( 1 ) for carrying out a method according to one of the preceding claims, characterized by an environment detection device ( 2 ) for detecting an environment of a vehicle (F) and objects located therein (O1, O2) and - an evaluation device ( 3 ) for evaluation by means of the environment detection device ( 2 ) detected environmental data (D), for determining in free-flowing traffic from the vehicle (F) accessible surface areas from the environmental data (D), for the prediction of trajectories (TF ₁ to TF _n , TO1 ₁ to TO1 _m , TO2 ₁ to TO2 _z ) of the vehicle (F) and the objects (O1, O2) from the environmental data (D), for classifying a behavior of the objects (O1, O2) and for determining a probability whether the vehicle (F) collision-free in the free areas the fluent traffic is feasible, taking into account a cooperative behavior of the objects (O1, O2) from the environmental data (D) and - one with the evaluation device ( 3 ) coupled warning device ( 4 ) for issuing a driver warning when a predetermined target value is undershot and / or - one with the evaluation device ( 3 ) coupled control device ( 5 ) for automatic longitudinal and / or lateral control of the vehicle (F) when falling below a predetermined desired probability.

Images