METHOD FOR OPERATING A MOTOR VEHICLE ASSISTANCE SYSTEM AND AN ASSISTANCE SYSTEM

Abstract

Technologies and techniques for operating an assistance system in a motor vehicle with which a driving a shoulder by the motor vehicle is detected by means of at least one detection device in the assistance system, and with which a control signal is generated by means of an electronic computing device in the assistance system for engaging with a functional unit in the motor vehicle. The shoulder that is detected is evaluated in terms of the type of shoulder, and a critical or non-critical driving of the vehicle on the shoulder is determined on the basis of the type of shoulder, and the control signal is generated in the case of a critical driving on the shoulder. The present disclosure also relates to an assistance system.

Claims

1-10. (canceled)

11. A system for operating an assistance system in a motor vehicle, comprising: a computing device; a functional unit, operatively coupled to the electronic device; and at least one detection device configured to visually detect a shoulder of a road, wherein the computing device is configured to process the detected shoulder of a road using machine learning to classify a type of shoulder, the shoulder type comprising one of a critical and non-critical shoulder, and wherein the computing device is configured to generate and transmit a control signal for the functional unit if a non-critical shoulder type is detected.

12. The system of claim 11, wherein the at least one detection device comprises at least one camera.

13. The system of claim 11, wherein the at least one detection device comprises at least two wheel sensors on different wheels of the motor vehicle, and wherein the computing device is configured to detect and classify a type of shoulder on the basis of data from at least two wheel sensors.

14. The system of claim 13, wherein the computing device is configured to process the detected shoulder of the road to classify the type of shoulder on the basis of a comparison between temporal courses of the at least two wheel speed sensors.

15. The system of claim 11, wherein the computing device is configured to detect and classify a type of shoulder on the basis of different image patterns received from the detection device.

16. The system of claim 11, wherein the detection device comprises at least two vertical motion sensors and/or two wheel speed sensors on opposite wheels on the motor vehicle.

17. The system of claim 11, wherein the computing device is configured to process the detected shoulder of the road to determine a road surface of the road, and wherein the computing device is configured to process the detected shoulder of the road to classify the type of shoulder utilizing the road surface.

18. A method for operating an assistance system in a motor vehicle, comprising: visually detecting, via at least one detection device, a shoulder of a road, processing, via a computing device, the detected shoulder of a road using machine learning to classify a type of shoulder, the shoulder type comprising one of a critical and non-critical shoulder, generating and transmitting a control signal for a functional unit of the assistance system if a non-critical shoulder type is detected.

19. The method of claim 18, wherein the at least one detection device comprises at least one camera.

20. The method of claim 18, wherein the at least one detection device comprises at least two wheel sensors on different wheels of the motor vehicle, and wherein the detecting and classifying a type of shoulder comprises detecting and classifying on the basis of data from at least two wheel sensors.

21. The method of claim 20, wherein processing the detected shoulder of the road to classify the type of shoulder comprises processing the detected shoulder on the basis of a comparison between temporal courses of the at least two wheel speed sensors.

22. The method of claim 18, wherein detecting and classify the type of shoulder comprises detecting and classify the type of shoulder on the basis of different image patterns received from the detection device.

23. The method of claim 18, wherein the detection device comprises at least two vertical motion sensors and/or two wheel speed sensors on opposite wheels on the motor vehicle.

24. The method of claim 18, further comprising processing the detected shoulder of the road to determine a road surface of the road, and wherein classifying the type of shoulder comprises utilizing the road surface.

25. A system for operating an assistance system in a motor vehicle, comprising: a computing device; a functional unit, operatively coupled to the electronic device; and at least one detection device configured to visually detect a shoulder of a road, wherein the computing device is configured to process the detected shoulder of a road using machine learning to classify a type of shoulder, the shoulder type comprising one of a critical and non-critical shoulder, wherein the computing device is configured to process the detected shoulder of the road to determine a road surface of the road, and wherein the computing device is configured to process the detected shoulder of the road to classify the type of shoulder utilizing the road surface, and wherein the computing device is configured to generate and transmit a control signal for the functional unit if a non-critical shoulder type is detected.

26. The system of claim 11, wherein the at least one detection device comprises at least one camera.

27. The system of claim 11, wherein the at least one detection device comprises at least two wheel sensors on different wheels of the motor vehicle, and wherein the computing device is configured to detect and classify a type of shoulder on the basis of data from at least two wheel sensors.

28. The system of claim 13, wherein the computing device is configured to process the detected shoulder of the road to classify the type of shoulder on the basis of a comparison between temporal courses of the at least two wheel speed sensors.

29. The system of claim 11, wherein the computing device is configured to detect and classify a type of shoulder on the basis of different image patterns received from the detection device.

30. The system of claim 11, wherein the detection device comprises at least two vertical motion sensors and/or two wheel speed sensors on opposite wheels on the motor vehicle.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Exemplary embodiments of the present disclosure shall be explained below in reference to the schematic drawings. Therein:

[0026] FIG. 1 shows a schematic top view of an embodiment of a motor vehicle that includes an assistance system, under some aspects of the present disclosure.

DETAILED DESCRIPTION

[0027] The exemplary embodiments described below are preferred exemplary embodiments of the present disclosure. The components described in the exemplary embodiments each represent individual, independent features of the present disclosure, each of which independently develop the present disclosure and are therefore to be regarded individually, or in combinations other than those shown, as components of the present disclosure. Furthermore, the exemplary embodiments described herein can also be supplemented by other features of the present disclosure described herein.

[0028] Functionally identical elements are provided with the same reference symbols in the drawing.

[0029] FIG. 1 shows a schematic top view of an embodiment of a motor vehicle 1 that has an embodiment of an assistance system 2. The assistance system 2 has an electronic computing device 3. The assistance system 2 also has at least one detection device 4, 5, 6. By way of example, the assistance system 2 can contain a camera 4 serving as the detection device 4, 5, 6. Furthermore, a respective wheel speed sensor 5 can form the detection device 4, 5, 6, and/or a respective vertical motion sensor 6 can form the detection device 4, 5, 6, on a respective wheel 7 of the motor vehicle 1.

[0030] Under this configuration, it is possible to detect when the motor vehicle 1 is driving 8 on the shoulder 9 of a road by means of the assistance system 2. In particular, the driving 8 on the shoulder 9 is detected by means of the at least one detection device 4, 5, 6, and a control signal 10 is generated by means of the electronic computing device 3 for engaging with a functional unit 11 in the motor vehicle 1 based on the detected shoulder driving 8.

[0031] The shoulder 9 of the road is evaluated with regard to a type of shoulder, and it is determined whether the motor vehicle is driving 8 on a critical or non-critical shoulder on the basis of the determined type of shoulder, and the control signal 10 is only generated in the case of a critical driving 8 on the shoulder.

[0032] In particular, the shoulder 9 of the road can be evaluated with regard to the type of shoulder by means of machine learning on the part of the electronic computing device 3.

[0033] A classification of whether the vehicle is driving 8 on the shoulder, and if so, which type of shoulder this is, takes place by means of the assistance system 2. By way of example, it is possible to determine that a shoulder, formed by grass, sand, gravel, etc. has a low friction coefficient, and can therefore be regarded as critical in particular. It is also possible to detect a shoulder with a high friction coefficient, made of cobblestone, paving stone, grass paving stone, grass plastic paving, or rumble strips, which is non-critical, in particular. The classification of the shoulder by pattern recognition takes place with an evaluation of the respective sensor data using a sensor-specific evaluation software. In particular, the detection devices 4, 5, 6 can be regarded as sensors in this case. The evaluation software is trained for the classification of the types of shoulders described above using learning data. Learning data comprise the temporal courses of the respective sensor data recorded during a test drive when travelling over the different types of shoulders specified above. The training of the evaluation software takes place using known processes, in particular machine learning.

[0034] This information, regarding whether a vehicle is driving 8 on a shoulder, and the type of shoulder, form the input data for a respective evaluation software for the respective sensors, or the respective detection devices 4, 5, 6, which are connected via a data bus to a data fusion algorithm in the electronic computing device 3. This data fusion algorithm determines from the input data obtained from the detection devices 4, 5, 6 whether the vehicle is driving 8 on the shoulder, and whether this is a critical shoulder with a low friction coefficient. If both are the case, the data fusion algorithm sends this information to a triggering algorithm. If a strong steering reaction by the driver in order to get off the shoulder 9 is detected at the same time, this triggering algorithm activates the shoulder function, represented in particular by the control signal 10. In particular, a selective braking of the wheels 7 on the shoulder 8 of the road can be carried out by means of an electronic stabilization system, wherein the electronic stabilization system (ESP) corresponds to the functional unit 11 in particular.

[0035] The determination of the type of shoulder can be carried out in particular on the basis of the different detection devices 4, 5, 6. By way of example, crossing the lane lines 12 or the edge of the roadway, corresponding in particular to the transition from asphalt to the shoulder 9, can be detected using the front camera 4. These data are already used in a so-called lane departure warning assistant or a lane keeping assistant, for example. Image sequences are recorded for this. An image data evaluation software is used for the learning, in that different typed of shoulders generate specific image patterns in the image sequences in the camera data, which can be assigned to the respective types of shoulders in order to distinguish them from one another. After the training, the image processing software can assign the data recorded with the front camera 4 while driving the vehicle on the different types of shoulders, or classify them, as described above.

[0036] Furthermore, it is possible to draw conclusions regarding the roughness of the substrate using the wheel speed sensors 5. These data can also serve in helping the anti-blocking system or a dynamic stability system, or to prevent rattling and humming in transmissions. These wheel speed sensors 5 record sensor data corresponding to a temporal course of the wheel speeds of the four wheels 7 on the motor vehicle 1. When driving over a rough surface, as may be the case when driving 8 on a shoulder, for example, the data for the temporal courses of these wheel speed sensors fluctuates substantially, even though the vehicle speed remains practically unchanged while driving 8 on the shoulder. The fact that the different types of shoulders generate specific patterns or specific fluctuations in the temporal courses of the wheel speed sensor data, which can be assigned to the respective types of shoulders, and thus distinguish them from one another, is exploited when training a wheel speed data evaluation software. Many of the shoulder surfaces with a high friction coefficient, e.g. rumble strips made of concrete slabs, lane marking lines 12, grass paving stone, etc. also exhibit uniform structures, which are reflected in the sensor data from the wheel speed sensors 5. After the training, the wheel speed data evaluation software can assign the data recorded with the wheel speed sensors 5 while driving the vehicle to the different types of shoulders, or classify them therein, as described above.

[0037] It is also possible to record the wheel movements in the vertical direction using vertical motion sensors 6, e.g., in a vertical regulation system in the motor vehicle 1, and therefore also draw conclusions regarding the roughness of the substrate therewith. The sensor data from the vertical motion sensors 6 relate to the temporal courses of the vertical motions of the four wheels 7 on the motor vehicle 1. On a rough substrate, the temporal courses of these sensor data fluctuate. The fact that the different types of shoulders generate specific patterns or specific fluctuations in the temporal courses of the vertical motion sensor data, which can be assigned to the respective types of shoulders, and therefore distinguish them from one another, is exploited in the training of the vertical motion data evaluation software. Many of the shoulder surfaces with a high friction coefficient, e.g., rumble strips made of concrete slabs, lane marking lines 12, grass paving stone, etc. also exhibit uniform structures which are reflected in the sensor data from the vertical motion sensors 6. After the training, the height data evaluation software can assign the data recorded with the vertical motion sensors 6 while driving the vehicle to the different types of shoulders, or classify them therein, as described above.

[0038] To determine whether or not the vehicle is driving 8 on the shoulder, the data from the wheel speed sensors 5 and/or the data from the vertical motion sensors 6 on the wheels 7 on the left and right are compared with one another. If the data from the wheel speed sensors 5 and/or the data from the vertical motion sensors 6 from the wheels 7 on both the left and the right indicate a rough substrate, then the motor vehicle 1 is very probably not driving on the shoulder 9, but instead on cobblestones, a dirt road, or some other poor road surface. If the data from the wheel speed sensors 5 and/or the data from the vertical motion sensors 6 for either the wheels 7 on the left or right indicate a rough substrate, the motor vehicle is potentially driving with either the left wheels 7 or the right wheels 7 on the shoulder 9. Whether or not this is critical is determined using the respective evaluation software, as described above.

[0039] Potentially critical special situations, such as a roadway 13 made of cobblestones and a shoulder 9 made of grass or sand, can be detected via simultaneous classification of the surfaces of the road 13 and the shoulder 9 using the method described above.

[0040] Specific structures on the road surface generated by reflections, shadows and grooves, for example, can likewise lead to an erroneous detection of driving 8 on the shoulder by the camera 4. Because learning data can also be generated for these specific structures on the road surface, the image data evaluation software from the camera 4 can be trained for this such that these specific structures are classified as non-critical. The erroneous detection of driving 8 on a shoulder in the case of specific structures on the road surface can thus be prevented.

[0041] Unevenness in the road surface, e.g., manhole covers, road damage, tar seams, corrugation in the asphalt, or Botts' dots in the asphalt, can likewise result in an erroneous detection of driving 8 on the shoulder by the wheel speed sensors 5. Because these types of unevenness also generate specific patterns in data, the wheel speed evaluation software and/or vertical motion evaluation software can be trained for this, such that they are classified as non-critical. The erroneous detection of driving 8 on the shoulder in the case of unevenness in the road surface can thus be prevented.

[0042] On the whole, the present disclosure results in a method for classifying types of shoulders.

LIST OF REFERENCE SYMBOLS

[0043] 1 motor vehicle [0044] 2 assistance system [0045] 3 electronic computing device [0046] 4 camera [0047] 5 wheel speed sensor [0048] 6 vertical motion sensor [0049] 7 wheel [0050] 8 driving on the shoulder [0051] 9 shoulder [0052] 10 control signal [0053] 11 functional unit [0054] 12 lane marking line [0055] 13 road