DE102018203376A1 - Method for detecting and taking into account irregular driving behavior of a target vehicle - Google Patents

Abstract

The invention relates to a method for detecting and taking into account irregular driving behavior of a target vehicle (2) with the following steps: detecting lane boundaries by means of at least one surrounding field sensor of an ego vehicle (1) and / or referring to the lane boundaries from a highly accurate map; Lane lane boundaries to a position of the target vehicle (2) from the detection of the target vehicle by the at least one surrounding detection sensor of the ego vehicle (1), - Determining whether an irregular driving behavior of the target vehicle (2) is present, the irregular driving behavior resulting from a behavior of the target vehicle (2) within the traffic lane limits and / or an airspeed and / or an acceleration and / or deceleration and / or an unnatural driving maneuver of the target vehicle (2) is derived, - at least one driving parameter of the ego vehicle (1) being anticipated after the determination of the irregular driving behavioradapted to reduce the potential risk with the target vehicle (2).

Description

The invention relates to a method for detecting and taking into account irregular driving behavior of a target vehicle.

From the prior art, for example, a method is known which describes an automatic detection of unsafe or dangerous vehicles. When such a vehicle is detected, a warning should be issued to the driver of the ego vehicle and possibly to other vehicles.

It is therefore an object of the present invention to further develop the method known from the prior art, so that the safety in road traffic, especially in mixed traffic, is increased.

This object is solved by the subject matter of independent claims 1 and 9.

Further advantageous embodiments and further developments are the subject of the dependent claims.

Initial considerations were that issuing a warning in automated driving is not effective or sufficient. Against this background, it is necessary to adapt the driving strategy of the ego vehicle in order to ensure traffic safety, especially in mixed traffic. Mixed traffic here describes a scenario in which manually controlled and automated vehicles meet. Furthermore, the number of different sensors in the vehicles is steadily increasing, so that several different sensor data can be used for the reliable detection of unsafe vehicles.

• - Erkennen von Fahrspurbegrenzungen mittels zumindest einem Umfelderfassungssensor eines Ego-Fahrzeugs und/oder Beziehen der Fahrspurbegrenzungen aus einer hochgenauen Karte,
• - Zuordnung der Fahrspurbegrenzungen zu einer Position des Zielfahrzeugs aus der Erkennung des Zielfahrzeugs durch den zumindest einen Umfelderfassungssensor des Egofahrzeugs,
• - Feststellen ob ein irreguläres Fahrverhalten des Zielfahrzeugs vorliegt, wobei das irreguläre Fahrverhalten aus einem Verhalten des Zielfahrzeugs innerhalb der Fahrspurbegrenzungen und/oder einer Eigengeschwindigkeit und/oder einer Beschleunigung und/oder Verzögerung und/oder eines unnatürlichen Fahrmanövers des Zielfahrzeugs abgeleitet wird,

wobei nach dem Feststellen des irregulären Fahrverhaltens vorausschauend zumindest ein Fahrparameter des Ego-Fahrzeugs zur Verringerung einer potentiellen Gefahr mit dem Zielfahrzeug angepasst wird.According to the invention, therefore, a method for detecting and taking into account irregular driving behavior of a target vehicle is proposed with the following steps:

• Detection of lane boundaries by means of at least one environment detection sensor of an ego vehicle and / or acquisition of lane boundaries from a highly accurate map,
• Assignment of the lane boundaries to a position of the target vehicle from the recognition of the target vehicle by the at least one surroundings detection sensor of the ego vehicle,
• Determining whether irregular driving behavior of the target vehicle exists, wherein the irregular driving behavior is derived from a behavior of the target vehicle within the lane boundaries and / or an airspeed and / or an acceleration and / or deceleration and / or an unnatural driving maneuver of the target vehicle,

wherein, after determining the irregular driving behavior, at least one driving parameter of the ego vehicle is anticipated to be adapted to reduce a potential danger with the target vehicle.

An environment detection sensor in this context is a camera system, a lidar sensor system or a radar sensor system. Preferably, a combination of two or all three of these environment detection sensors is used.

The term lane boundaries here describes lane boundary lines or sublime boundaries such as crash barriers, edge structures or construction site boundaries. By means of a lidar sensor system or a camera system, both the lane boundary lines and the aforementioned raised lane boundaries and thus also a lane course can be detected. By means of a radar sensor system, the raised lane boundaries can be detected and thus the lane course can be estimated at least and accordingly also an irregular driving behavior of the target vehicle can be detected by means of radar.

Cumulatively or alternatively for detecting the target vehicle by means of a radar sensor system and / or a camera system and / or Lidarsensorsystems the ego vehicle, it would also be possible that the target vehicle independently information about the position and / or the trajectory trajectory to the ego vehicle by car-to Car communication passes. Thus, the ego vehicle would use the information received to determine if irregular driving is present.

A potential hazard here may be, for example, an unexpected emergency braking of the target vehicle to a standstill, a collision with another road user or with a lane boundary or the like.

The behavior within the lane encompasses whether the target vehicle has a constant trajectory within the lane boundaries or whether lateral deviations can be detected. This means that it is determined by means of at least one of the sensors, whether the target vehicle within the lane boundaries snaking lines or exceeds the lane boundaries several times in succession. If such a behavior is detected, the driving behavior of the target vehicle is classified as irregular. A constant Trajectory describes a trajectory that runs approximately centrally within two lane boundaries. A constant trajectory can alternatively also be present when the target vehicle is moving on the basis of a reference trajectory, which does not conclude an irregular driving behavior in the event of deviations from the lane center due to lane constrictions and / or obstacles.

Alternatively or cumulatively, the intrinsic speed and / or an acceleration and / or deceleration and / or an unnatural driving maneuver of the target vehicle can also be monitored to determine the irregular driving behavior. By monitoring the airspeed or the acceleration and / or deceleration of the target vehicle, for example, it can be ascertained whether the target vehicle is driving slower or faster than allowed for no apparent reason, such as dense traffic or bad weather conditions such as smoothness, fog or the like cause suddenly slows down or accelerates sharply. An excessive speed or a strong acceleration can be caused, for example, by driving prohibited races, which represents a very great danger, especially in urban areas. The maximum speeds or typical speeds can be stored in a map. Dense traffic can be detected by means of on-board sensor technology or on-line traffic information. Smoothness through on-line weather information or on-board temperature and rain sensor or on-board road condition detection.

An unnatural driving maneuver here describes, for example, a sudden turning of a target vehicle or a sudden lane change, for example, away from a turn lane on a straight track.

The target vehicle may in this case be a manually controlled vehicle, which is controlled by an impaired driver. For example, the driver of this vehicle may be affected by alcohol, drugs, fatigue, or other influences. Furthermore, the target vehicle may be a vehicle driven by a non-resident driver. A non-resident driver is unlikely to have an accurate knowledge of the location's roads. This leads to an increased risk of unnatural driving maneuvers or sudden braking, so a strong deceleration of the target vehicle, as they tend to inquire more often for the way or miss junctions. However, the target vehicle can also be an automated vehicle with a malfunction, in which case, for example, one or more sensors are disturbed, which can lead to an irregular driving behavior.

The danger space describes the environment of the target vehicle and extends, for example, to the respectively adjacent lanes and a rear space behind the target vehicle. The extent in the rear space of the target vehicle can be determined, for example, by the minimum safety distance of the ego vehicle to the target vehicle. The minimum safety distance depends on the current speed of the ego vehicle and can be set and maintained independently by the ego vehicle. Thus, the distance between the first vehicle and the target vehicle would be completely classified as a danger area. The lateral boundaries of the danger space can be determined for example by the field of view of a front camera and / or by the detection range of a lidar and / or radar sensor.

In a preferred embodiment, the adaptation of a driving parameter comprises the adaptation of the speed of the ego vehicle and / or the adaptation of a lateral and / or longitudinal distance to the target vehicle. Such an adaptation is advantageous in order to reduce the potential danger, since, for example due to a lower speed and / or an increased distance more time for evasion and / or deceleration remains before a direct risk of collision and an emergency braking maneuver of the ego vehicle should be performed. Preferably, the longitudinal distance is increased by the reduction in speed. The increase of the lateral distance is possible, for example, on a three-lane road.

For example, if the target vehicle is on the far right lane, it may be switched to the leftmost lane.

Furthermore, all further road users located in the danger area are preferably classified as potentially unsafe. This means that it is assumed that these other road users can also potentially have an irregular driving behavior. In such a scenario, the safety distance is increased to the target vehicle and another road user to dodge or decelerate, for example, in a collision of the target vehicle with the other road users.

Particularly preferably, the ego vehicle is operated fully automatically or semi-automatically. This means that the ego vehicle is equipped with appropriate control and computing units, which are designed to the vehicle without Participation of the driver situational control. In an emergency, such as a sensor failure, it may still be possible for the driver to take control of the ego vehicle.

In a further preferred embodiment, a trajectory of the ego vehicle is planned in order to avoid a collision with the target vehicle, to shield the danger space or to leave the danger space. For shielding the danger space, the ego vehicle can, for example, arrange centrally between two lanes so that a following vehicle can not get into the danger area. Alternatively, the ego vehicle may orient itself toward an emergency lane to more quickly avoid an abrupt deceleration of the target vehicle. Furthermore, it would be possible, for example, in a three-lane highway to overtake the target vehicle on the leftmost lane when the target vehicle is on the far right lane, if there is no other road users on the middle lane and thus a sufficient lateral safety distance between the Ego vehicle and the target vehicle is present.

In the planning of the trajectory, it is particularly preferable to take into account a speed of the target vehicle, a width of the roadway, a number of lanes on the roadway, and a number of other road users.

In a particularly preferred embodiment, an information is output to other road users and / or an adjacent infrastructure after determining the irregular driving behavior of the target vehicle. A simple way of alerting other road users would be to turn on the hazard warning lights. Alternatively, the location of the target vehicle or information about the danger area can be transmitted to subsequent vehicles via car-to-car communication, for example. Furthermore, it would be possible via Car to X communication to pass the information about the target vehicle to the adjacent infrastructure, for example, to issue a warning on motorways on scoreboards to other road users or regulate the maximum permissible speed via the scoreboards, which the speed of the ego Vehicle so that subsequent vehicles reduce the speed early and so do not get near the danger area.

According to the invention, a system for carrying out the proposed method is furthermore provided.

Preferably, the system comprises at least one lidar sensor and / or a camera and / or a radar sensor and a computing unit for evaluating the sensor data to detect lane boundaries and / or a target vehicle and to plan a trajectory of the ego vehicle.

Particularly preferably, the system comprises a control unit in order to adapt at least one driving parameter of the ego vehicle.

Further advantageous embodiments will be apparent from the drawings.

• 1: ein schematisches Ablaufdiagramm einer Ausführungsform des Verfahrens;
• 2: eine schematische Darstellung einer Ausführungsform der Erfindung;
• 3: eine weitere schematische Darstellung einer weiteren bevorzugten Ausführungsform der Erfindung.

Show:

• 1 Fig. 3 is a schematic flow diagram of one embodiment of the method;
• 2 a schematic representation of an embodiment of the invention;
• 3 : a further schematic representation of another preferred embodiment of the invention.

1 shows a schematic flow diagram of an embodiment of the method. In step S1 lane boundaries, for example, by a lidar sensor of the ego vehicle 1 detected. In a next step S2 becomes the lane boundary of a position of a target vehicle 2 assigned. The target vehicle 2 can be detected by lidar, radar and / or camera. In step S3 then becomes the behavior of the target vehicle 2 monitored and evaluated within lane boundaries to determine if irregular driving is present. If an irregular driving behavior is detected, then in a next step S4 at least one driving parameter of the ego vehicle 1 customized. According to a preferred embodiment, now in a further step S5 the environment around the target vehicle 2 as a danger area G be classified if an irregular driving behavior has been detected.

2 shows a schematic representation of an embodiment of the invention. Here is a schematic illustration of a side of a freeway with two lanes FS1 . FS2 with the same direction of travel R shown. Furthermore, there is an emergency lane NS or a so-called streak strip available. On the lane FS1 is the ego vehicle 1 in the direction of travel R1 as well as the target vehicle 2 in the direction of travel R2 , At the time t ₁ is the ego vehicle 1 at a first position 1t ₁ and detects a position 2t ₁ of the target vehicle 2 which deviates from an expected position because they are not centrally located within the lane FS1 located. Now, the behavior of the target vehicle will become 2 within the lane FS1 observed. At the time t ₂ is the ego vehicle 1 at a position 1t ₂ , The target vehicle 2 is at this time t ₂ at a position 2t ₂ , which is a lateral deviation from the position 2t ₁ having. At a time t ₃ is the ego vehicle 1 at a third position 1t ₃ , The target vehicle 2 is in time t ₃ at a position 2t ₃ showing a lateral deviation both to the first position 2t ₁ as well as the second position 2t ₂ having.

Due to these different positions of the target vehicle 2 within the lane FS1 becomes the driving behavior of the target vehicle 2 regarded as irregular and the target vehicle 2 classified as unsafe.

In 3 a further schematic representation of a further preferred embodiment of the invention is shown. Here is how in 2 a schematic representation of a side of a highway with two lanes FS1 . FS2 with the same direction of travel R shown. Here is the ego vehicle 1 and the target vehicle 2 one behind the other on the lane FS1 , On the adjacent lane FS2 there is another road user 3 right next to the target vehicle 2 , The reference numerals R1 . R2 and R3 each describe the direction of travel of the ego vehicle 1 , the target vehicle 2 and the other road user 3 and are in the same direction R directed.

The ego vehicle 1 has the target vehicle in this case 2 classified as unsafe according to the proposed method due to irregular driving behavior. The environment around the target vehicle 2 is then used as a danger area G classified. The other road user 3 is initially in the danger area G and is thus classified as unsafe. The ego vehicle 1 Now, for example, the distance to the target vehicle 2 and the other road user 3 increase in order to avoid or brake in the event of a possible collision. This is done by reducing the speed of the ego vehicle 1 reached. Alternatively or cumulatively, the ego vehicle can 1 also change his trajectory. For example, the ego vehicle 1 a trajectory Ta plan, whereby the ego vehicle between the lanes FS1 and FS2 located to the danger area G for future vehicles. Another possibility would be a trajectory Tb , which makes the ego vehicle the emergency lane NS approaches to dodge faster on this in an abrupt deceleration of the target vehicle.

LIST OF REFERENCE NUMBERS

1

Ego vehicle

2

target vehicle

3

other road users

FS1

first lane

FS2

second lane

G

danger area

NS

emergency lane

R

direction of travel

R1

Driving direction Ego vehicle

R2

Driving direction of the target vehicle

R3

Driving direction of other road users

S1-S5

steps

t ₁ -t ₃

timings

Ta

first possible trajectory

Tb

second possible trajectory

Claims (11)

Method for detecting and taking into account irregular driving behavior of a target vehicle (2) with the steps: Detection of lane boundaries by means of at least one environment detection sensor of an ego vehicle (1) and / or obtaining lane boundaries from a highly accurate map, Assignment of the lane boundaries to a position of the target vehicle (2) from the recognition of the target vehicle by the at least one surroundings detection sensor of the ego vehicle (1), - Determining whether an irregular driving behavior of the target vehicle (2) is present, wherein the irregular driving behavior of a behavior of the target vehicle (2) within the lane boundaries and / or an airspeed and / or acceleration and / or deceleration and / or an unnatural driving maneuver of the target vehicle (2) is derived - After the determination of the irregular driving behavior, at least one driving parameter of the ego vehicle (1) is anticipated to be adapted with the target vehicle (2) in order to reduce a potential danger. Method according to Claim 1 wherein the adaptation of a driving parameter comprises the adaptation of the speed of the ego vehicle (1) and / or the adaptation of a lateral and / or longitudinal distance to the target vehicle (2). Method according to Claim 1 , wherein after determining the irregular driving behavior, the environment of the target vehicle (2) is classified as a danger area (G). Method according to one of the preceding claims, wherein all further road users (3) located in the danger area (G) are classified as potentially unsafe. Method according to one of the preceding claims, wherein the ego vehicle (1) is operated fully automatically or semi-automatically. Method according to one of the preceding claims, wherein a trajectory (Ta, Tb) of the ego vehicle is planned in order to avoid a collision with the target vehicle in order to shield the danger space (G) or around the danger space (G) of the target vehicle (2). to leave. Method according to Claim 6 , wherein in the planning of the trajectory (Ta, Tb), a speed of the target vehicle (2), a width of the road, a number of lanes (FS1, FS2) on the road and a number of other road users (3) is taken into account. Method according to one of the preceding claims, wherein after determining the irregular driving behavior of the target vehicle (2) information to other road users (3) and / or a surrounding infrastructure is issued. System for carrying out a method according to the Claims 1 to 8th , System after Claim 9 wherein the system comprises at least one lidar sensor and / or a camera and / or a radar sensor and a computing unit for evaluating the sensor data to detect lane boundaries and / or a target vehicle (2) to detect an irregular driving behavior of the target vehicle (2) and to plan a trajectory (Ta, Tb) of the ego vehicle (1). System according to one of Claims 9 or 10 wherein the system comprises a control unit to adapt at least one driving parameter of the ego vehicle (1).

Images