DE102014204107A1 - Traffic forecasting method - Google Patents

Abstract

A method for forecasting a traffic space available for an ego vehicle (1) comprises the following steps: - at least two foreign vehicles (2, 3) are detected, - states of the foreign vehicles (2, 3) are categorized, - movements of the foreign vehicles (2, 3) are predicted - in the case of forecasting a collision of the foreign vehicles (2, 3), a blocking area (SP) to be avoided by the ego vehicle (1) is determined, which depends on the categorization of the foreign vehicles (2, 3).

Description

The invention relates to a method for forecasting a traffic space available for a vehicle. Furthermore, the invention relates to a system for carrying out such a prognosis method.

From the DE 10 2012 006 986 A1 For example, a method for traffic space prognosis is known, which should be applicable in particular to autonomous vehicles, but also to a vehicle controlled by a human driver. The vehicle using the traffic forecasting method is always referred to as an ego vehicle. To distinguish from this vehicle other vehicles that use the same traffic space, referred to as foreign vehicles. The method predicts the movements of other vehicles and predicts possible interference between the predicted movements, which suggests that an evasive maneuver of another vehicle is imminent. An evasive maneuver is considered to be a deviation from the predicted movement and leads to the definition of restricted zones which are not available for the ego vehicle within the traffic area. As part of the traffic space prediction method, it can be provided that the driver of the ego vehicle receives an indication, for example in the form of an acoustic warning signal, about the available traffic space. If the ego vehicle is an autonomous vehicle, an intervention in its steering and / or speed control can also be provided.

The DE 102 44 205 A1 discloses a method for preventing the collision of vehicles, in which a hazard level is determined from a right of wrong approach and a Kollisionsgefährdungsungsmaß. Depending on the risk level determined, various steps can be taken to reduce the risk of collision. The steps may in particular include interventions in the brake control of the vehicle.

The invention has for its object to further develop a method for traffic control over the above-mentioned prior art with regard to specific reaction options to different traffic situations.

This object is achieved by a method for predicting an available traffic space with the features of claim 1 and by a trained for performing such a forecasting system according to claim 7. The following explained in connection with the method embodiments and advantages of the invention apply mutatis mutandis to the device, that is, the traffic forecasting system, and vice versa.

The procedure assumes a traffic situation with at least three vehicles. This is the designated as ego vehicle, which performs the traffic space forecast, as well as at least two other vehicles, also referred to as foreign vehicles. The technical equipment with which the method is carried out is at least partially, preferably completely, in or on the ego vehicle. The traffic forecasting system thus represents a vehicle-based system. The ego vehicle is a motorized, single or multi-lane vehicle. The foreign vehicles may be any motorized or non-motorized vehicles.

The first step of the method for traffic space prognosis is the detection of at least two foreign vehicles by the vehicle-based sensor technology of the ego vehicle. The sensor technology of the ego vehicle can here be based on any physical principles. Examples include optical systems, radar and lidar systems. For near-field sensors, ultrasound-assisted systems are also possible. Essential for the further implementation of the method is the categorization of states of the foreign vehicles by the vehicle-based forecasting system. A state of a vehicle is understood to mean both its state of motion and its design. In simple embodiments of the method, the categorization refers either exclusively to the state of movement of the foreign vehicles or exclusively to the design of the foreign vehicles. In refined variants of the prognosis method, both the state of movement and the type of another vehicle are included in its respective categorization.

In all cases, movements of the foreign vehicles are predicted, and in the case of forecasting a collision between the foreign vehicles, a restricted area to be avoided by the ego vehicle is determined. The geometry of this restricted area depends on the categorization of the other vehicles.

Certain categories of states of the foreign vehicles are advantageously assigned to different types of crash scenarios. Different crash scenarios can be differentiated, for example, with regard to the angle at which the predicted trajectories of the two foreign vehicles meet. To define different crash scenarios, different categories of vehicle types can be formed, which For example, differ from each other in terms of vehicle mass and / or vehicle dimensions.

As a crash scenario I, for example, a collision between two foreign vehicles is defined, which touch each other laterally, that is to say travel substantially in the same direction, and have only a slight difference in speed relative to one another. In this type of collision, it is to be assumed that the two foreign vehicles involved in the collision essentially continue to move in their original direction of travel without slipping. Due to the collision itself and / or knee-jerk reactions of the drivers involved in the foreign vehicles, however, a clear sideways component of the directions of movement of the foreign vehicles after the collision is to be expected. Depending on the parameters of the collision, in particular the absolute and differential speeds, it may be most sensible for the driver of the ego vehicle to use a traffic area which is clearly off the traffic area used by the foreign vehicles, or specifically to control that area of the traffic area in which the collision takes place, since the foreign vehicles are likely to move very quickly laterally from this area. The traffic areas to be avoided by the driver of the ego vehicle are referred to as restricted areas.

For example, a second crash scenario, which has typical differences from crash scenario I, concerns a collision between a first, relatively light vehicle, namely passenger cars, and a second, a multiple heavier vehicle, namely, a truck, which was originally much slower than that first foreign vehicle, that is the car, drives. Next is the crash scenario II a significant lateral offset of the two foreign vehicles given to each other, so that the car as the first foreign vehicle with only a part of his front on the front of him in the same direction moving truck impinges. In contrast to the crash scenario I, in this case, it can be assumed that the car immediately skids as a result of the collision, rapidly reduces speed, and thereby moves strongly to the side, relative to the original direction of travel. The result is an exclusion zone which is much larger than the exclusion zone in the case of the first crash scenario.

In a preferred embodiment, the system also checks whether, in the case of the prognosis of a collision of the foreign vehicles, as a result of this collision, a free driving lane for the ego vehicle in the area of the current lanes of the foreign vehicles will result. This is based on the consideration that in some selected crash scenarios, the collision itself gives the vehicles a strong impulse to opposite sides, thereby releasing between them a space that may provide a free route for the ego vehicle. In this case, therefore, the crash parameters are suitably checked and evaluated.

Furthermore, in a preferred embodiment, a risk assessment is carried out for different options, in particular for different driving lanes for the ego vehicle, and in particular those with the lowest risk potential for the ego vehicle are selected and proposed or automatically adopted for the vehicle driver.

In principle, different measures can be triggered in the ego vehicle in the different crash scenarios by determining the restricted zones and free driving lanes. Is the risk that the ego vehicle is involved in the accident, relatively low and the available response time, in particular due to a large distance to the location of the collision between the foreign vehicles for the driver of the ego vehicle sufficiently long, so can be sufficient to the driver of the ego vehicle directed visual and / or audible warning. If, on the other hand, there is a high risk of the collision of the ego vehicle with one of the other vehicles, an automatic intervention in the steering of the ego vehicle may be provided. At the highest risk for the ego vehicle further measures known per se can be taken automatically, such as the automatic actuation of the brakes, the erection of the seat backs of the ego vehicle, the closing of a possible sunroof and the activation of belt tensioners.

Several embodiments of the invention will be explained in more detail with reference to a drawing. Herein show:

1 a first collision event between two foreign vehicles observed by an ego vehicle,

2 in a representation analog 1 a second collision event,

3 in a table the categorization of different crash scenarios.

The 1 illustrates the collision event referred to above as Crash Scenario I. An ego vehicle is with 1 , two foreign vehicles with 2 . 3 designated. The vehicles 1 . 2 . 3 are on a highway 4 on the way, which three lanes 5 . 6 . 7 in the same direction. This drives the ego vehicle 1 in the left lane 5 , the first foreign vehicle 2 on the middle lane 6 , and the second foreign vehicle 3 in the right lane 7 , The ego vehicle 1 is located at some distance, in 1 not shown to scale, behind the foreign vehicles 2 . 3 and drives a little faster than the other vehicles 2 . 3 , Between the foreign vehicles 2 . 3 there is only a small difference in speed.

In a collision between the foreign vehicles 2 . 3 These touch laterally over much of their length. An unillustrated sensor within the ego vehicle 1 , which in particular comprises a radar sensor system and / or optical sensors, supplies to a likewise in the ego vehicle 1 located evaluation unit, which is another component of the vehicle-based forecasting system, data, which - starting from the in 1 sketched traffic situation - on the imminent collision between the foreign vehicles 2 . 3 close. Driving directions of other vehicles 2 . 3 after the predicted collision are in 1 indicated by dashed arrows. Between these dashed arrows is a blocking surface SP, that of the ego vehicle 1 to avoid. How out 1 As can be seen, the blocking surface SP is initially completely within the lanes, as viewed from the location of the collision 6 . 7 , Has the ego vehicle 1 a sufficient excess of speed over the foreign vehicles 2 . 3 , so driving on the left lane 5 the best measure to being a driver of the ego vehicle 1 not to be involved in an accident yourself. A corresponding signal is given to the driver within the ego vehicle 1 , For example, in the form of a straight arrow on a display, displayed. An intervention in the steering, brake or engine control of the ego vehicle 1 is not intended.

This in 2 sketched collision event corresponds to the already described as crash scenario II scenario. It resembles the crash scenario I insofar as the ego vehicle 1 on the three-lane highway 4 behind the foreign vehicles 2 . 3 on the left lane 5 moves. In contrast to crash scenario I, the second foreign vehicle is involved 3 but not a passenger car, but a truck. The categorization of the second foreign vehicle 3 as a truck is made by the built in the ego vehicle I forecasting system. For the categorization of foreign vehicles 2 . 3 In addition to the sensors already mentioned, acoustic sensors can also be used in individual cases, depending on the type of vehicle and the traffic situation.

In the in 2 Outlined traffic situation is essentially on the middle lane 6 , partly in the right lane 7 moving foreign vehicle 2 to a passenger car, which far faster than the second foreign vehicle 3 that means the truck drives. The inevitable, through the forecasting system in the ego vehicle 1 detected in advance collision between the foreign vehicles 2 . 3 causes the first foreign vehicle 2 a strong burst to the left towards the left lane 5 receives. This is in 2 by a dashed arrow, which from the first foreign vehicle 2 goes out and the locking surface SP limited to one side, indicated.

In crash scenario II, it can not only be assumed that the first foreign vehicle 2 after the collision with the second foreign vehicle 3 on the left lane 5 is hurled, but also that the first foreign vehicle 2 this quickly loses speed. Accordingly, in the ego vehicle 1 Measures are taken as quickly as possible to avoid entering the restricted zone SP. These measures can in particular an immediate, automatic braking intervention after detecting the in 2 presented traffic situation, even before the collision between the foreign vehicles 2 . 3 , lock in.

If, due to the existing crash parameters, for example in the crash scenario I or the crash scenario II, it can be assumed that one or both foreign vehicles 2 . 3 If, for example, the system gains a strong impulse, the system checks and determines, if appropriate, in a preferred embodiment, whether in this case the point of the collision and / or the previous lane 6 of the vehicle 2 under certain circumstances even a possible option as a driving route so no restricted area SP for the ego vehicle 1 is. In this case, it is therefore to be expected that one or both of the colliding vehicles 2 . 3 will move away to the side. If such an option is recognized, it will be the driver of the ego vehicle, possibly considering a risk to other options for driving lanes 1 signaled or the driving route automatically taken by the vehicle itself.

In 3 By way of example, some possibilities of classifying crash scenarios are summarized in tabular form. Different types of vehicle collisions FK can be found in individual columns of the table. In a first type of vehicle collision, referred to in the table as FK1, is a collision between two cars, as in FIG 1 shown. Another type of vehicle collision, referred to in the table as FK2, is a collision between a car and a truck, as in FIG 2 shown. In the individual rows of the table, crash scenarios are distinguished according to the angle at which the other vehicles involved in the collision 2 . 3 collide. A first angle range W1 summarizes scenarios in which the two foreign vehicles 2 . 3 originally in the same or almost the same Driving direction, as in both the example 1 as well as in the example 2 the case is.

In contrast, collision scenarios are summarized under the angle range W2, in which the two other vehicles 2 . 3 at an angle significantly greater than zero, for example, an angle of 45 ° to 90 °, collide. Further, not shown in the figures collision scenarios, for example, head-on collisions between similar or different types of foreign vehicles 2 . 3 affect. The in the table after 3 Exemplary crash scenarios are referred to as C1 / 1, C1 / 2, C2 / 1 and C2 / 2, whereby the crash scenarios according to the 1 and 2 about the crash scenarios C1 / 1 and C1 / 2. In every crash scenario C1 / 1, C1 / 2, C2 / 1 and C2 / 2 will be in the ego vehicle 1 automatically triggered a customized response, which in particular depends on the type of foreign vehicles involved 2 . 3 depends.

LIST OF REFERENCE NUMBERS

1

Ego vehicle

2

foreign vehicle

3

foreign vehicle

4

Highway

5

left lane

6

middle lane

7

right lane

C1 / 1

Crash scenario

C1 / 2

Crash scenario

C2 / 1

Crash scenario

C2 / 2

Crash scenario

FK1

vehicle collision

FK2

vehicle collision

SP

exclusion zone

W1

angle range

W2

angle range

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

• DE 102012006986 A1 [0002]
• DE 10244205 A1 [0003]

Claims (9)

Method for predicting an ego vehicle ( 1 available traffic area, comprising the following steps: - at least two third-party vehicles ( 2 . 3 ) are recorded, - States of other vehicles ( 2 . 3 ) are categorized, - movements of other vehicles ( 2 . 3 ) are forecasted, - in the case of forecasting a collision of the foreign vehicles ( 2 . 3 ) becomes one of the ego vehicle ( 1 ) prohibited area to be avoided (SP) determines which of the categorization of the other vehicles ( 2 . 3 ) depends. Method according to Claim 1, characterized in that the categorization of the states of the external vehicles ( 2 . 3 ) corresponds to different types of crash scenarios (C1 / 1, C1 / 2, C2 / 1, C2 / 2). A method according to claim 2, characterized in that the different types of crash scenarios (C1 / 1, C1 / 2, C2 / 1, C2 / 2) in terms of the angle, below the trajectories of other vehicles ( 2 . 3 ), to be distinguished from each other. Method according to one of claims 1 to 3, characterized in that the foreign vehicles ( 2 . 3 ) are categorized according to vehicle types. A method according to claim 4, characterized in that the foreign vehicles ( 2 . 3 ) are categorized by weight classes. Method according to one of claims 1 to 5, characterized in that it is checked whether in the case of forecasting a collision of the foreign vehicles ( 2 . 3 ), as a result of the collision, a free road for the ego vehicle ( 1 ) in the area of the current driving lanes of other vehicles ( 2 . 3 ). Method according to one of claims 1 to 6, characterized in that in the case of forecasting a collision of the foreign vehicles ( 2 . 3 ) an automatic intervention in the control of the ego vehicle which depends on its categorization ( 1 ) he follows.  System for detecting and evaluating traffic situations, designed to carry out the method according to claim 1. System according to claim 8, characterized in that this is designed as a vehicle-based system.

Images