CN111051817B

CN111051817B - Method, control device and computing device for determining the position of a motor vehicle

Info

Publication number: CN111051817B
Application number: CN201880051101.8A
Authority: CN
Inventors: F·施威泽
Original assignee: Bayerische Motoren Werke AG
Current assignee: Bayerische Motoren Werke AG
Priority date: 2017-09-14
Filing date: 2018-09-11
Publication date: 2023-09-12
Anticipated expiration: 2038-09-11
Also published as: CN111051817A; DE102017216238A1; WO2019053013A1

Abstract

The invention relates to a method for determining the position of a motor vehicle (10) in an environment, wherein at least one landmark (15) in the environment is provided with a corresponding position specification (16) of the landmark (15), and the motor vehicle (10) is identified as being correspondingly currently passing by the at least one landmark (15), and the position specification (16) of the corresponding currently passed landmark (15) is output as the position of the motor vehicle (10). Provision is made here for a driving maneuver characteristic which is generated when the respective landmark (15) is passed, which gives a predetermined unique characteristic (30) of the respective driving maneuver, and for the description data of the respective driving maneuver currently being driven by the motor vehicle (10) to then be generated by means of at least one detection device of the motor vehicle (10), for the predetermined characteristic (20) of the driving maneuver to be extracted from the description data, and for the currently passed landmark (15) to be identified by comparing the extracted characteristic (20) with the unique characteristic (30) in the respective maneuver characteristic of the at least one landmark (15).

Description

Method, control device and computing device for determining the position of a motor vehicle

Technical Field

The invention relates to a method for determining the position of a motor vehicle in the environment. For this purpose, the motor vehicle is known to pass a predetermined landmark, the position of which is known. The position of the motor vehicle can thus be deduced from the position of the landmark. A control device for a motor vehicle for carrying out the method according to the invention also belongs to the invention. Furthermore, a stationary computing device is provided by the present invention.

Background

A method of the above-mentioned type is known, for example, from DE102011112404 A1. According to the known method, an object in the environment is detected as a landmark by means of a camera and the relative position of the motor vehicle with respect to the object is then determined, in order to determine the position of the motor vehicle itself from the known position of the object and the relative position. Such an approach requires costly object recognition in the camera image data. Furthermore, it must be possible to measure the relative position of the motor vehicle with respect to the object.

A method is known from DE102013016435B4, by means of which a deceleration strip (bodenshwellen) on a road surface is identified by a motor vehicle. Since the position of the speed bump is known, the position of the motor vehicle can be regarded as the same as the position of the speed bump when driving over the speed bump. The method identifies a deceleration strip by means of vibration patterns, such as a booming sound in a motor vehicle when driving over the deceleration strip. But this requires an analysis of the costs of the solid-state acoustic sensor signal

Disclosure of Invention

The invention is based on the task of determining or ascertaining the position of a motor vehicle in the environment using simple technical means.

This object is achieved by the solution according to the invention. Further advantages result from the embodiments presented in the following description and in the drawings.

The invention relates to a method for determining the position of a motor vehicle in an environment, wherein at least one landmark of the environment is provided with a corresponding position specification of the landmark, and the motor vehicle is identified as passing by the at least one landmark accordingly, and the position specification of the landmark which is passed by the motor vehicle is output as the position of the motor vehicle, characterized in that: -providing a manipulation characteristic of a driving maneuver generated when passing a respective landmark, wherein the manipulation characteristic of the driving maneuver comprises a sequence of driving maneuver steps predetermined when passing a respective landmark, wherein the respective manipulation characteristic gives a predetermined unique feature of the respective driving maneuver, and-generating description data of the driving maneuver currently being traveled by the motor vehicle by means of at least one detection device of the motor vehicle, and extracting the predetermined feature of the driving maneuver from the description data, and-identifying the landmark currently being passed by comparing the extracted feature with the unique feature in the respective manipulation characteristic of the at least one landmark.

The invention provides a method for determining the position of a motor vehicle in an environment. The position may be an absolute geographical position or a relative position, for example a relative position with respect to a predetermined reference object. The method uses solutions known from the prior art, i.e. at least one predetermined landmark for the environment is provided with a corresponding position specification of the landmark. That is, at least one landmark and its location in the environment are known. The location of the landmark is described by a corresponding location specification. According to the method, the motor vehicle is correspondingly currently passing the at least one landmark is identified. That is, during each trip, it can be recognized when the motor vehicle has traveled past or past a known landmark. The position specification of the respective currently traversed landmark is then output as the respective current position of the motor vehicle.

The invention relates to how such landmarks can be identified by technically simple means. For this purpose, the steering characteristics of the driving maneuver which occur when the respective landmark is passed are provided. In particular, driving maneuvers are understood to be a chronological sequence of steering and/or acceleration and/or braking adjustments. For example, the landmark may be a unique double curve (Doppelkurve) of the road, which double curve may be identified by means of the corresponding steering behavior of the motor vehicle. The respective manipulation characteristic of the landmark describes a predetermined characteristic or unique feature of the respective driving manipulation, such as first turning right and then turning left. In other words, the steering feature describes how the motor vehicle must travel or guide when passing the landmark. I.e. the landmarks are identified by means of longitudinal guidance (acceleration and/or braking) and/or lateral guidance (steering) of the motor vehicle. Accordingly, according to the invention, the description data of the driving maneuver currently being driven by the motor vehicle is generated by means of at least one detection device of the motor vehicle. Then, a predetermined feature of the travel maneuver is extracted from the description data. The currently traversed landmark is identified or recognized by comparing the extracted features of the current driving maneuver with the unique features in the corresponding maneuver characteristic of the at least one landmark. For this purpose, a predetermined consistency criterion can be defined for the features to be compared. For example, a predetermined tolerance interval may be defined for differences between features, within which the consistency of the respective feature is always indicated. Then, if all the extracted features or at least a predetermined minimum portion of the extracted features are consistent with the respective unique features of the manipulation characteristic in accordance with the consistency criterion, the landmark described by the manipulation characteristic is identified as the one currently traversed.

The invention has the advantage that only the driving behavior of the motor vehicle itself, in particular with respect to longitudinal and transverse guidance, has to be detected in order to derive or determine the position of the motor vehicle therefrom. It is not necessary to detect objects (visible objects and/or deceleration strips) outside the vehicle in a costly manner.

The invention includes further embodiments by which additional advantages are derived.

In one embodiment, the at least one detection device is used to generate description data based on an estimated voyage (koppelenaliation) based on the movement of the vehicle of the motor vehicle. I.e. by means of the odometer (Odometrie) of the motor vehicle. In other words, the detection of the vehicle movement is not dependent on position signals of GNSS (Global Navigation Satellite System ), such as GPS (Global Positioning System, global positioning system). Such GNSS position signals have a larger variance or dispersion than position detection by means of an odometer. That is, the vehicle movement can thus be determined more precisely than by means of the position signals of GNSS.

According to one embodiment, the at least one detection device is used to determine description data about at least one of the following driving dynamics variables: wheel speed of at least one wheel, acceleration in at least one spatial direction, yaw rate, travel speed, steering angle. The at least one detection device can be provided with a sensor and/or a signal processing device for this purpose. As the sensor, for example, a wheel speed sensor, an acceleration sensor, a steering angle sensor, and/or a yaw rate sensor may be provided. The driving speed can be determined by means of a speed detection device, which can be based on, for example, the measured wheel speed of the vehicle wheels. That is, the description data may include sensor data and/or processed and/or combined sensor data. The described driving dynamics variables can be detected reliably in the motor vehicle. The detection means may also be formed by the above-described receiver of the position signals of the GNSS, but this receiver will only be used in combination with at least one further detection means here.

In particular, one embodiment provides that the recognition of the passing landmarks is carried out independently of the image recognition of the objects in the environment. Thereby advantageously not depending on, for example, visibility in the environment.

According to one embodiment, the description data are detected by means of a plurality of detection devices. In order to combine the description data from the different detection devices at this time, it is provided that the description data are combined by means of a motion model of the motor vehicle. This ensures that a reliable description of the driving maneuver is obtained. An example of such a motion model is the single-track model (einpurmoldel), known per se.

One embodiment provides that the unique features of the predetermined handling characteristic and the features extracted from the description data respectively comprise at least one of the following features: a marked entry point for a landmark, a marked exit point for a landmark, a radius of a driving curve, a curvature of a driving curve, a yaw rate, a direction of entry toward the landmark (e.g., one azimuth), a direction of exit from the landmark (e.g., one azimuth), a curve turning point. The curve turning point can be obtained by extending the driving-in direction and the driving-out direction, for example, and the intersection point of the driving-in direction and the driving-out direction is the curve turning point. The entry direction extends here from the entry point of the curve and the exit direction extends from the exit point of the curve. The described features have proven to be advantageous in reliably repeatedly identifying landmarks.

An embodiment provides that the at least one landmark respectively comprises at least one intersection and/or at least one curve. The corresponding driving maneuvers during the driving through at least one intersection and/or curve have advantageously proven to be reliably identifiable repeatedly. For this purpose, additionally or alternatively, a cornering maneuver and/or a cornering maneuver may be provided as a unique driving maneuver for the respective landmark, respectively. The sequence of steering and/or acceleration and/or braking adjustments produced here yields features that can be reliably repeatedly detected.

According to one embodiment, it is provided that the individual driving characteristics are provided for different lanes or traffic lanes of the surrounding road (in particular for adjacent traffic lanes). The position determination can thus be carried out either specifically to the lane line or specifically to the traffic lane.

The more landmarks are identified, the greater the number of comparisons that need to be made to compare features extracted from the current driving maneuver with the maneuver characteristics of the landmarks. In order to limit the number, an embodiment provides that the temporary geographic position of the motor vehicle is determined by means of a receiver for the position signals of the GNSS. However, in the described manner, the temporary geographic position thus determined is accompanied by a discrepancy or variance and is therefore accordingly imprecise. At least one operating characteristic is selected from a plurality of stored operating characteristics, the associated position of which in a predetermined environmental range is located around the temporary geographic position. That is, only checking which landmarks are likely to be located only nearby by means of the temporary geographic location. The comparison of the extracted features of the current driving maneuver is then limited to the selected at least one maneuver characteristic. The computational effort for carrying out the method can thus be limited by specifying the size of the environment.

In this case, according to one embodiment, it can be provided that the environmental area has a diameter of more than 10m, in particular more than 20 m. In particular, the diameter is preferably greater than the dispersion specific to the receiver of the position signals of the GNSS. This advantageously ensures that the correct landmark is in the set of selected landmarks. In order to limit the computational effort, it is preferably provided that the environmental area has a diameter of less than 500m, in particular less than 300 m.

As already explained, the driving maneuver is preferably determined on the basis of an odometer of the motor vehicle. The odometer may have drift and/or offset. It is therefore preferably provided that the deviation of the position detection based on the odometer, which may be caused, for example, by drift and/or offset, is determined or corrected by means of the respective position of the motor vehicle determined at the at least one landmark. That is, in particular, deviations of the trajectory detection based on the relative position detection are corrected or at least reduced. Thus, advantageously, odometer-based position detection can also be used for long travel path (in particular longer than 500 m) travel path detection.

Some embodiments relate to the question of how the respective steering characteristics of the respective landmarks can be formed. The method steps associated therewith constitute a self-aspect of the invention, which can be carried out even without the aforementioned method steps. In order to determine the respective driving characteristics of the at least one landmark, description data describing the respective driving maneuver of the other vehicle when the respective landmark is passed are preferably received from a plurality of other vehicles, respectively. The other vehicle may be a vehicle that has passed the landmark once itself before the vehicle that is currently passing the landmark. Corresponding predetermined features are extracted based on description data of other vehicles. A respective frequency distribution is formed from the extracted features and then a respective one of the unique features is derived from the respective frequency distribution. This gives the advantage that the travel path of the individual further vehicle, which is displaced relative to the center of the roadway, is detected as a statistically anomalous measurement. That is, the unique feature may be, for example, an average of extracted features from other vehicles. This gives the advantage that the influence of the respective travel path of each individual further vehicle is relative. Examples for such unique features are: average curve radius, average entry point to landmark, average exit point from landmark, average absolute driving direction (Heading), for example as azimuth specification.

In one embodiment, two landmarks specific to the traffic lane are identified in at least one frequency distribution by means of mathematical local extrema (i.e., maxima and/or minima) of the frequency distribution. For distinguishing between two traffic lanes, a predetermined traffic lane width may be used for reliability.

In order to carry out the method, the invention also provides a control device for a motor vehicle. The control device may be provided in the motor vehicle, for example, as at least one controller or coupled to the motor vehicle via a radio connection as a server of the internet. As a server, the control device may be provided by a stationary computing device, i.e. as a computer or computer network. A hybrid form with at least one controller and at least one server may also be provided. The control device is provided for carrying out an embodiment of the method according to the invention.

The invention also relates to the described formation of the respective steering characteristics of the at least one landmark. For this purpose, a stationary computing device for operation on a data network is provided, wherein the computing device is provided for carrying out the method having the method steps described above in connection with the formation of the operating characteristics.

Other features of the invention will be apparent from the accompanying drawings and from the description of the drawings. The features and feature combinations mentioned above in the description and those mentioned in the following description of the figures and/or shown individually in the figures can be used not only in the respectively given combination but also in other combinations or alone.

Drawings

The invention will now be described in more detail by means of preferred embodiments and with reference to the accompanying drawings. Wherein:

fig. 1 shows a schematic illustration of a motor vehicle whose position is detected by an embodiment of the method according to the invention;

fig. 2 shows a diagram for elucidating the method steps of an embodiment of the method according to the invention, which method can be carried out by a control device;

FIG. 3 shows a graph illustrating a statistical distribution of features of a driving maneuver along a landmark;

FIG. 4 shows a diagram illustrating unique features specific to detecting the steering characteristics of two landmarks to a roadway; and

fig. 5 shows a graph for elucidating the statistical distribution of the unique features of the steering characteristic of fig. 4, by means of which a distinction can be made between two traffic lanes.

Elements having the same function are correspondingly provided with the same reference numerals in the figures.

Detailed Description

Fig. 1 shows a motor vehicle 10. The motor vehicle 10 may be, for example, a car or a van. The motor vehicle 10 can start from the past or historical position 11, travel along the road 12, through the travel path 13, and the travel path is also determined by means of an estimated voyage. The travel path 13 can be determined by means of an estimated travel method as a relative position change over time, which is produced by travel along the travel path 13. However, the dead reckoning method used may have a deviation or drift 14 that needs to be compensated for. In motor vehicle 10, a predetermined landmark 15 is thus identified during travel along road 12, whose position 16 can be known in geographic coordinate system 17 and can be described by a corresponding position specification. The corresponding position 16 is synonymously denoted here with the associated position specification. Such a landmark 15 may accordingly be, for example, a curve or a turn possibility of the road 12. In other words, the motor vehicle 10 must perform a predetermined sequence of actuation steps (i.e., steering and/or acceleration and/or braking operations) specific to the landmark 15 when passing or driving over the respective landmark 15. These can be determined by means of an odometer of the motor vehicle 10.

With the aid of the receiver for the GNSS position signals, the current position of the motor vehicle 10 can also be detected accordingly in the motor vehicle 10, but only with an accuracy that is less than the accuracy of the dead reckoning. In other words, the determined travel path 13 of the reference relative position change is more accurate than the receiver for GNSS, but the absolute position of the travel path 13 of the determined reference geographic coordinate system 17 is subject to the described drift 14, for example.

In order to eliminate or compensate for the drift 14, a corresponding driving-over landmark 15 is detected in the motor vehicle 10, and the determined driving trajectory 13 is then arranged or determined in a coordinate system 17, in particular to the driving lane, by means of an offset 18, based on a position specification of the corresponding position 16 of the detected landmark 15. The method steps carried out for this purpose can be carried out by a control unit ECU of the motor vehicle 10, for example by a controller.

In order to identify the landmark 15, the vehicle 10 can first determine the area or environment 19 in which the vehicle 10 is currently located by means of a receiver for the GNSS position signals. The GNSS position determined for this purpose constitutes a rough position fix or a temporary geographical position, i.e. a position fix with accuracy for the receiver of the position signals. The temporary geographic location may constitute a center point of the environmental scope 19. It is then possible, for example, to determine in the database which possible landmarks 15 are currently likely to be in the vicinity of the motor vehicle 10. All landmarks 15 whose locations 16 are within the current region 19 are selected. Thereafter, the motor vehicle 10 can correspondingly extract the predetermined features 20 from the description data of the detection device during some driving maneuvers of the motor vehicle 10, for example, a curve driving maneuver or a cornering maneuver. In fig. 1, curve entry points 21, curve entry directions 22, curve exit points 23, curve exit directions 24, curve turning points 25 and curve radii 26 are illustrated as extracted features 20 for curve driving. The curve entry point 22 and the curve exit point 23 can be identified, for example, by means of the steering wheel or the temporal change in the steering angle position of the wheel. In detecting the features 20, the influence of the drift 14 is insignificant on the basis of the relatively small spatial extension of the respective landmark 15 (for example less than 1km, in particular less than 500 m) and thus the respective landmark 15 can be repeatedly identified by means of the extracted features 20.

At this time, for each landmark 15, an average value for the corresponding feature 20 may be stored for each of the evaluated or extracted features 20. Such averages are examples for typical or unique features. The entirety or collection of unique features of the landmarks 15 constitute the steering characteristics of driving over the respective landmark 15. Thus, for all landmarks 15 that are within the current region 19, the extracted features 20 may be compared to the unique features of the respective landmark 15, i.e., to the steering features of the landmark 15. Here, a tolerance interval may be set in order to confirm the consistency of the extracted feature 20 with the corresponding unique feature (e.g., for the feature "curve radius").

If the currently extracted feature 20 is identified as consistent with the steering characteristics of the landmark 15, the landmark 15 is signaled as the current driven landmark. The position 16 of the currently traveled or passed landmark 15 is then used as the current position of the motor vehicle 10 in order to determine the determined travel path 13 by means of the offset 18 at the known position 16 in the manner described.

Fig. 2 illustrates how unique features for individual landmarks 15 can be found. The method steps described below may be implemented by a stationary computing device SRV, which may be implemented as a server of the internet, for example. The computing device SRV may communicate with the motor vehicle via a corresponding communication connection for data exchange. The corresponding communication connection may comprise, for example, a mobile radio connection and/or an internet connection.

The unique characteristic is formed in this example as an average of predetermined characteristics which are respectively extracted for a plurality of vehicles from the driving maneuver of the vehicle when passing or driving over the respective landmark 15. For illustration, the landmark 15 of fig. 2 may be a turn possibility 27, such as an intersection. The plurality of vehicles have passed the landmark 15 over the time of the vehicle 10. The plurality of motor vehicles is referred to below for better differentiation as other vehicles. And more particularly to selected vehicles or survey vehicles whose location is known. From which the travel tracks 28 can be received respectively. In general, it may be provided that each other vehicle, when passing through the landmark 15, finds the features 29 extracted for the corresponding driving maneuver performed when passing through the landmark 15. Next, as an example, a curve radius K as an example for such extracted features 29 shall be further discussed. Fig. 2 illustrates how one of the other vehicles can determine the radius value K1 as an extracted feature 29 of the curve radius K.

Fig. 3 illustrates how a curve radius K (generally as an example for the extracted feature 29) when plotting the radius values for all other vehicles for the landmark 15 (i.e. here the turn possibility 27) yields a frequency distribution H as a frequency map in which the radius value K1 represents a possible value. Based on the frequency distribution H, a unique feature 30 can then be determined or ascertained, for example, as the highest frequency value of the feature 29 (here, i.e., the curve radius K, for example). And can thus be processed with all extracted features 29. The thus-obtained set of unique features 30 then yields the steering characteristics of the landmark 5.

Fig. 4 illustrates how the distinction can be made by unique features specifically to the traffic lane. Fig. 4 illustrates a cornering situation 31 similar to fig. 2, but in which two traffic lanes 32 can be driven. Each lane 32 constitutes its own landmark.

Fig. 4 and 5 together illustrate how the frequency distribution H can be derived from the extracted features 29 of the driving trajectories 28 of all other vehicles, which frequency distribution can have two local maxima 33 for one extracted feature 29 (for example, the curve radius K), which suggests a two-lane road. For example, a smaller curve radius K1 is always produced in the inner lane, while a larger curve radius K2 is always produced for the outer lane. From the frequency distribution H, the local maxima 33 can be used again, for example, as the respective unique features 30 of the two landmarks 15 for the cornering possibility 31.

A particularly preferred embodiment is again described below. The accuracy of the navigation data or the trajectory data of the driving trajectory can be subdivided into global and local areas. The object of global navigation, e.g. GNSS, is a coarse positioning with an accuracy of a few meters. However, this accuracy is not sufficient for reliably distributing the vehicle onto the traffic lane. In local navigation, it is possible to assign a position specific to the traffic lane, for which purpose however reference points, so-called landmarks 15, must be present. Heretofore, the generation and identification of landmarks has only been possible visually with camera systems or via LIDAR.

Alternatively, it is proposed to use a sensor device or a general detection device, which is installed in the motor vehicle in mass production (i.e. a receiver for the GNSS position signals and additionally at least one wheel speed sensor, acceleration sensor, yaw rate sensor and/or speed measurement). If the combined results of the sensors are taken into account, the unique landmarks 15 can be identified.

For this purpose, repeated driving maneuvers are detected in the road network and the maneuvering characteristics are established via a plurality of driving passes. A cornering manoeuvre and a curve with at least one predetermined minimum curvature are suitable for this.

The travel path is derived with a smaller variance from a combination of a plurality of detection devices than if the GNSS position sequence were considered alone. The association may be implemented, for example, based on a motion model (e.g., a single-track model) of the motor vehicle. In this case, the curvature of the roadway can be estimated, in particular, from consideration of the wheel speed and the yaw rate. It is noted that not every driver drives through the selected curve in the middle of the driving lane, but selects his own driving line. Therefore, the curvature of the actual running has a frequency distribution over a plurality of running passes (fig. 3 and 5). By this distribution, the curve curvature characteristic as a unique characteristic can be normalized. Thus, for example, for each curve, an average suspension point is also produced at the curve input and the curve output

In the area of intersections, the curve maneuvers are differentiated by means of different features, for example, the calculated curve radius, the direction of entry and the direction of exit, and are each classified as a curve feature. In the case of a multi-lane cornering possibility 31, different steering characteristics and thus also individual suspension points (landmarks) for the respective cornering lane can be produced in this way. Fig. 2 and 4 show the generation of landmarks 15 for a right curve. The relevant travel track has been previously determined via the entry direction to the intersection and the exit direction from the intersection. The average curve radius can be determined from the frequency maps of a plurality of driving curves (fig. 3 and 5). In the case of a right turn of a multilane (fig. 4), the driving direction is no longer sufficient for a map specific to the traffic lane. A plurality of peaks or local maxima 31 (fig. 5) are generated in the frequency map, which describe the average curve radii of the different traffic lanes 32.

The advantage of this solution is that the suspension points (landmarks) are produced by the sensor system of the motor vehicle. The descriptive data is present in an amount sufficient for accurate determination or repeated identification of the suspension point.

List of reference numerals

10. Motor vehicle

11. Historical location

12. Road

13. The determined driving track

14. Drift of

15. Landmark

16. Position of landmark

17. Coordinate system

18. Offset of

19. Region(s)

20. Extracted features

21. Curve entry point

22. Direction of travel

23. Curve driving out point

24. Direction of travel

25. Bend turning point

26. Radius of curve

27. Possibility of turning

28. Travel track

29. Extracted features

30. Unique features

31. Possibility of turning

32. Traffic lane

33. Local maxima

H frequency distribution

Radius of K curve

K1 Radius value

K2 Radius value

Claims

1. Method for determining the position of a motor vehicle (10) in an environment, wherein at least one landmark (15) of the environment is provided with a respective position specification (16) of the landmark (15), and wherein the motor vehicle (10) is recognized as accordingly currently passing the at least one landmark (15), and the position specification (16) of the correspondingly currently passed landmark (15) is output as the position of the motor vehicle (10), characterized in that:

-providing a steering characteristic of the driving maneuver generated when passing the respective landmark (15), wherein the steering characteristic of the driving maneuver comprises a sequence of predetermined driving maneuver steps when passing the respective landmark (15), wherein the respective steering characteristic gives a predetermined unique characteristic (30) of the respective driving maneuver, and

-generating description data of a driving maneuver currently being driven by the motor vehicle (10) by means of at least one detection device of the motor vehicle (10), and extracting predetermined features (20) of the driving maneuver from the description data, and

identifying the currently passed landmark (15) by comparing the extracted feature (20) with the unique feature (30) in the corresponding steering characteristic of the at least one landmark (15) and outputting a position specification of the identified landmark (15) as the position of the motor vehicle (10),

wherein a temporary geographic position of the motor vehicle (10) is determined by means of a receiver of the position signal for the GNSS and at least one operating characteristic is selected from a plurality of stored operating characteristics, the associated position of the at least one operating characteristic being located around the temporary geographic position in a predetermined environment range (19), and the comparison of the extracted characteristics (20) is limited to the selected at least one operating characteristic.

2. The method according to claim 1, wherein the description data is generated by means of the at least one detection device on the basis of an odometer based on the movement of the vehicle.

3. The method according to claim 1, wherein the description data about at least one of the following measurement variables is determined by means of the at least one detection device: wheel speed of at least one wheel, acceleration in at least one spatial direction, yaw rate, travel speed, steering angle.

4. A method according to any one of claims 1 to 3, wherein the identification of passing landmarks (15) is performed independently of the image identification of objects in the environment.

5. A method according to any one of claims 1 to 3, wherein the description data are determined by means of a plurality of detection devices and combined by means of a motion model of the motor vehicle (10).

6. A method according to any one of claims 1 to 3, wherein the unique features (30) and the features (20) extracted from the descriptive data respectively comprise at least one of the following features: a marked entry point (21), a marked exit point (23), a radius (26) of a driving curve, a curvature of a driving curve, a yaw rate, a driving-in direction (22), a driving-out direction (24), and a curve turning point (25).

7. A method according to any one of claims 1 to 3, wherein the at least one landmark (15) represents an intersection and/or a curve, respectively, and/or a turning maneuver and/or a curve travel is provided as a unique travel maneuver for the respective landmark (15).

8. A method according to any one of claims 1 to 3, wherein the steering characteristics of the individual traffic lanes (32) for one road of the environment are provided separately.

9. The method according to claim 1, wherein the environmental range (19) has a diameter of more than 10 m.

10. A method according to any one of claims 1 to 3, wherein deviations of the trajectory detection based on the relative position detection are corrected by means of the respective position of the motor vehicle (10) determined at the at least one landmark (15).

11. A method according to any one of claims 1 to 3, wherein, for ascertaining a respective steering characteristic of the at least one landmark (15) from a plurality of other vehicles, description data describing a respective driving maneuver of the other vehicles when passing the landmark (15) are received accordingly, and a respective predetermined feature (29) is extracted based on the description data, and a respective frequency distribution is formed from the extracted features (29), and one of the unique features (30) is ascertained from the respective frequency distribution.

12. Method according to claim 11, wherein two lane-specific landmarks (15) are identified in at least one frequency distribution by means of a mathematical local extremum (33) of the frequency distribution.

13. The method according to claim 9, wherein the environmental range (19) has a diameter of more than 20 m.

14. Control device for a motor vehicle (10), wherein the control device is provided for carrying out the method according to any one of claims 1 to 13.

15. A stationary computing device for running on a data network, wherein the computing device is arranged to implement the method of any of claims 1 to 13.