Method and apparatus for robustly positioning a vehicle

Abstract

A method for robustly locating a vehicle with the aid of a positioning system (10) is proposed. Sensor data of at least one sensor of the vehicle are provided, the sensor data containing information about the surroundings of the vehicle and / or its own movement. A digital map of the surroundings of the vehicle is provided, and a scene of the surroundings of the vehicle is determined based on the sensor data and / or the digital map. The robustness of the positioning system is derived from the determined scene, and the current robustness of the positioning system is matched based on the robustness derived from the scene. Based on the matched robustness, the sensor data are assigned to the digital map with the aid of an assignment module (18) of the positioning system. Based on the sensor data assigned to the digital map, the position and / or orientation of the vehicle is determined with the aid of a position determination module (20) of the positioning system. As a result, the computational overhead for the positioning can be reduced in an advantageous manner.

Description

Technical Field

[0001] The invention relates to a method for robustly localizing a vehicle, in particular an at least partially automated vehicle. The invention also relates to a localization system for robustly localizing a vehicle. Background Art

[0002] For the purpose of at least partially automated driving of vehicles, robust localization of vehicles is increasingly being developed. This enables the use of additional information from digital maps and sensor data from one or more sensors of the vehicle in a vehicle-local model of the surroundings. The latter can form the basis for planning driving behavior of the vehicle system (e.g. planning speed, deceleration and / or steering movements). The planning of driving behavior and / or the associated behavior decisions can be used as the basis for trajectory planning and control.

[0003] Typically, within the scope of robust positioning, sensor data from a plurality of different sensors for detecting the surroundings of the vehicle are combined with one another. For example, camera data from one or more cameras (optionally in conjunction with lane recognition), radar data from one or more radar sensors, lidar data from one or more lidar sensors, and / or any other data containing information about the surroundings of the vehicle are combined and / or fused.

[0004] In particular, for the purpose of geo-referencing, at least part of the sensor data and / or at least part of the combined sensor data can be assigned to a digital map. This is also referred to as map matching.

[0005] Therefore, methods for robust localization can be very expensive and computationally intensive. This can make it difficult to implement corresponding localization systems in vehicles. Summary of the invention

[0006] By means of embodiments of the present invention, an improved method for robustly localizing a vehicle and / or an improved localization system can be provided in an advantageous manner. In the method according to the present invention, the computational effort can be reduced in an advantageous manner in particular.

[0007] One aspect of the invention relates to a method for robustly locating a vehicle by means of a positioning system. The method comprises the steps of providing, determining and / or detecting sensor data of at least one sensor of the vehicle, wherein the sensor data contain information about the surroundings and / or the vehicle's own movement. In addition, the method comprises the step of providing a digital map of the surroundings of the vehicle. The method is characterized in particular by the following steps:

[0008] - determining a scene of the vehicle's surroundings based on sensor data and / or based on a digital map;

[0009] - deriving the robustness of the positioning system from the scenario sought (e.g. the required robustness);

[0010] - Adapting the current robustness of the positioning system based on the robustness derived from the scenario;

[0011] - assigning at least a portion of the sensor data to at least a portion of the digital map by means of an assignment module of the positioning system based on the matched robustness;

[0012] Based on the sensor data assigned to the digital map, the position and / or orientation of the vehicle is ascertained with the aid of a position determination module of the positioning system.

[0013] The vehicle may in particular be an at least partially automated vehicle. The method may be designated as a method for operating a positioning system and / or a method for controlling an at least partially automated vehicle.

[0014] The sensor data may be any data of one or more sensors for detecting the surroundings of the vehicle. The sensor data may be, for example, camera data of one or more cameras, radar data of one or more radar sensors, ultrasonic data of one or more ultrasonic sensors, lidar data of one or more lidar sensors, and / or any other data. The sensor data may also represent data composed of a combination and / or fusion of different data. Therefore, the method may include a step of combining data of different sensors in order to generate the sensor data.

[0015] The positioning system may refer to a positioning system inside the vehicle. The positioning system may be implemented, for example, in a control device of the vehicle. Alternatively or additionally, the positioning system may be implemented at least partially on a data processing device (e.g., a server and / or a cloud server) outside the vehicle. The vehicle and the data processing device may have an interface for communication and / or data exchange.

[0016] In general, the ascertained scene may represent a representation, an image and / or an estimation of at least a part of the surroundings of the vehicle and / or of at least one feature of the surroundings.

[0017] The ability of a positioning system to find the correct positioning of a vehicle (i.e., the correct position and / or orientation of a vehicle) under adverse ambient conditions and / or in an adverse ambient environment can be referred to as the robustness (or required robustness) of the positioning and / or positioning system. The degree of robustness (and / or robustness) can generally represent a measure of the correctness of the vehicle positioning and / or the probability of correctness. The positioning system can be configured to achieve the current degree of robustness and / or the currently predetermined degree of robustness during positioning. Therefore, the degree of robustness can represent the positioning requirements for the positioning system. Such positioning requirements can represent and / or describe the quality, quality and / or measure of the positioning correctness. Here, correctness can be defined by the accuracy of the positioning of the positioning system relative to the estimation of the digital map.

[0018] Advantageously, the positioning system can be adapted to the scenario by determining the scenario and adapting the current robustness based on the robustness derived from the scenario. This in turn can lead to a significant reduction in the computational overhead required for positioning. On the one hand, positioning can thereby be configured more quickly and more efficiently, and on the other hand, more computing power is available for other systems of the vehicle. As a result, energy consumption can also be reduced in an advantageous manner. Moreover, due to the reduced computational overhead, the requirements for hardware (e.g. control devices) can also be reduced, so that more cost-effective hardware (e.g. positioning systems) can be used for positioning.

[0019] In particular, the present invention can be considered to be based on the knowledge described below. A high robustness positioning of a vehicle can be associated with considerable computational overhead. Therefore, it may be necessary to find a compromise between computational overhead or computing power, positioning accuracy and robustness. In conventional systems for positioning, currently only camera data (optionally in combination with lane recognition), lidar data and (if necessary) other data are fused with one another. Optionally, outliers in the sensor data can be identified and removed in conventional systems.

[0020] Unlike conventional systems, according to the invention, the robustness and / or robustness of the positioning system are matched according to the scenario and the positioning requirements for positioning associated with the scenario. For example, near an intersection where there are other vehicles with potential high speeds, the positioning requirements are increased. In such a scenario, the robustness (and / or at least one positioning requirement) can be correspondingly increased and / or matched by means of the method according to the invention. In contrast, the operation of a vehicle in a parking lot may be associated with a low positioning requirement. In such a scenario, the robustness (and / or at least one positioning requirement) can also be correspondingly matched and / or reduced in an advantageous manner by means of the method according to the invention.

[0021] According to one embodiment of the invention, the step of adapting the current robustness comprises the following steps: adapting at least one parameter of an allocation module of the positioning system based on the robustness derived from the scenario. For example, the allocation module may be reparameterized.

[0022] According to one specific embodiment of the invention, at least one parameter of the allocation module describes the number of objects considered in the surroundings of the vehicle and / or the maximum distance to the vehicle up to which the sensor data are evaluated. In other words, the number of objects in the surroundings and / or the field of view of the surroundings can be adapted based on the derived robustness. For example, certain objects (such as trees, houses and / or other static or semi-static objects) can be hidden in the surroundings in order to adapt the current robustness based on the derived robustness. Alternatively or additionally, objects that are more than the maximum distance away can be hidden and / or removed.

[0023] According to one embodiment of the present invention, the matching of at least one parameter of the allocation module comprises at least one of the following partial steps:

[0024] - selecting map data of the digital map to allocate at least a part of the sensor data;

[0025] - Select the resolution of the digital map;

[0026] - selection of an algorithm for assigning the sensor data to the digital map (in particular a map matching algorithm);

[0027] - adapting at least one parameter of an algorithm for assigning sensor data to a digital map, in particular a map matching algorithm;

[0028] -Adapt the number of particles in the particle filter of the allocation module.

[0029] In order to match the current robustness, for example, only a portion of the information about the surroundings contained in the digital map can be used. In order to match the current robustness, a certain map type can also be obtained and / or used. In addition, a robust algorithm or an unrobust algorithm can be selected depending on the derived robustness. A robust algorithm can be based on, for example, spectral registration and / or matching of a cost function. An unrobust algorithm can be based on, for example, a point-to-point assignment and / or a distance metric within the scope of an iterative process. Alternatively or additionally, in order to match the current robustness, one or more parameters of the algorithm can be matched and / or changed. For example, the maximum distance of points forming corresponding pairs can be changed in an iterative algorithm for determining the closest point (iterative closest point, ICP algorithm). Alternatively or additionally, the sensor data can be filtered with the aid of a particle filter, and the number of particles can be matched to match the current robustness.

[0030] According to one embodiment of the present invention, the step of matching the current robustness comprises: matching the data processing of the sensor data. For example, the resolution (especially the time resolution) of the sensor data can be changed. The sensor data can also be smoothed, aggregated and / or preprocessed according to the robustness. In order to match the robustness, part of the sensor data can also be removed, and only other parts of the sensor data can be allocated to the digital map.

[0031] According to one embodiment of the present invention, the scene obtained includes: the covering and / or covering degree of a part of the vehicle's surroundings by objects in the surroundings. Alternatively or additionally, the scene includes the number of objects in the vehicle's surroundings. The scene may also include road types, such as highways, national roads, high-traffic roads, low-traffic roads, single-lane roads and / or multi-lane roads. In addition, the scene may include intersections, intersection types, parking lots and / or residential areas. The scene may also include the driving degree of the road, the number of other traffic participants in the vehicle's surroundings, the speed of the vehicle, the speed of other traffic participants and / or the number of obstacles in the vehicle's surroundings.

[0032] According to one embodiment of the present invention, the step of determining the scenario includes determining the degree of interference and / or consistency of the sensor data. Alternatively or additionally, the step of determining the scenario includes determining the validity date of at least a portion of the digital map. In other words, the consistency and / or deviation between the sensor data can be taken into account for determining the scenario. It can also be checked whether the digital map is up-to-date or outdated.

[0033] According to one embodiment of the invention, the method further comprises a step of generating a surroundings model of the vehicle based on at least a portion of the sensor data and / or based on at least a portion of the digital map, wherein the degree of robustness is derived based on the generated surroundings model. For example, the surroundings model can be generated based on the sensor data (optionally in combination with surroundings information in the digital map). The model of the surroundings of the vehicle can in particular enable and / or represent an estimation of the actual surroundings in which the vehicle is located, so that a scene can be estimated and / or determined with high accuracy based on the surroundings model.

[0034] According to one embodiment of the invention, the digital map has a georeferenced description of a recommended degree of robustness, wherein the degree of robustness is derived from the georeferenced description. In other words, the digital map may contain a degree of robustness recommended according to the geographic coordinates, based on which a degree of robustness for matching the current degree of robustness of the positioning system may be determined and / or derived.

[0035] According to one embodiment of the present invention, the method further comprises the step of determining scene changes and the step of deriving another robustness level based on the determined scene changes. Thus, the robustness level can be dynamically adapted in an advantageous manner when the scene and / or the vehicle's surroundings change.

[0036] According to one embodiment of the invention, the method further comprises the step of generating a control command for controlling the vehicle based on the determined position and / or orientation. The control command can then be transmitted to a control device of the vehicle, which can instruct the vehicle to implement the control command. The control command can include, for example, a speed predetermination, a speed, an acceleration, a deceleration and / or a steering movement.

[0037] Another aspect of the present invention relates to a positioning system for robustly positioning a vehicle, wherein the positioning system is configured to perform the method described above and below. The positioning system may have a data memory and a processor. A program element may be stored on the data memory, and when the program element is run on the processor, the program element instructs the positioning system to perform the method described above and below. At least one digital map and / or sensor data may also be stored in the data memory. The positioning system may also have an allocation module and / or a position determination module, which is used to allocate sensor data to a digital map, and the position determination module is used to obtain the positioning (i.e., position and / or orientation) of the vehicle. The positioning system may be installed in a vehicle and / or implemented in a vehicle. Alternatively or additionally, the positioning system may be implemented at least partially on a server.

[0038] Features, elements and / or steps of the method as described above and below may be features and / or elements of the positioning system and vice versa. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0040] Figure 1 A positioning system according to an embodiment of the present invention is shown;

[0041] Figure 2 A flow chart for explaining the steps of a method for robustly localizing a vehicle according to an exemplary embodiment of the present invention is shown;

[0042] Figure 3 A flow chart is shown for illustrating the steps of a method for robustly localizing a vehicle according to an exemplary embodiment of the present invention.

[0043] These drawings are only schematic and not to scale. In the drawings, identical, identically acting or similar elements are provided with the same reference numerals. DETAILED DESCRIPTION

[0044] Figure 1 A positioning system 10 according to one embodiment of the present invention is shown.

[0045] The positioning system 10 has a data memory 12 and a processor 14. At least one digital map and sensor data of at least one sensor of the vehicle are stored in the data memory 12. At least one digital map is also stored in the data memory 12.

[0046] Optionally, the positioning system 10 has a positioning device 16 for determining the position of the vehicle. The positioning device 16 can have a GPS sensor, for example.

[0047] The localization system 10 also has an assignment module 18 for assigning the sensor data to the digital map. The localization system 10 also has a position determination module 20 for determining the position (ie, the position and orientation) of the vehicle.

[0048] The positioning system 10 can be installed in the vehicle and / or implemented in the vehicle. Alternatively or additionally, the positioning system 10 can be implemented at least partially on a server, and data can be exchanged via an interface of the vehicle and the positioning system 10, such as control commands for controlling the vehicle and / or data describing the positioning of the vehicle.

[0049] Figure 2 A flow chart for explaining the steps of a method for robustly locating a vehicle using a locating device 10 according to an exemplary embodiment of the present invention is shown. The locating device 10 may correspond to the reference Figure 1 A positioning device 10 is described.

[0050] In step S1 , the method comprises providing sensor data of at least one sensor of the vehicle, wherein the sensor data contain information about the surroundings of the vehicle. Furthermore, step S1 comprises providing a digital map of the surroundings of the vehicle.

[0051] In step S2, a scene of the vehicle's surroundings is determined based on sensor data and / or based on a digital map. The scene of the vehicle's surroundings may include information about the following: the covering of a part of the vehicle's surroundings by objects in the surroundings, the number of objects in the vehicle's surroundings, the type of road, intersection, type of intersection, parking lot, residential area, the degree of travel on the road, the number of other traffic participants in the vehicle's surroundings and / or the number of obstacles in the vehicle's surroundings. In addition, the scene may include information about the speed of the vehicle and / or the speed of other traffic participants.

[0052] To determine the scenario in step S2 , the degree of interference and / or degree of consistency of the sensor data can optionally also be determined. Alternatively or additionally, a validity date of at least a portion of the digital map can be determined.

[0053] Furthermore, in step S2, a surrounding environment model of the vehicle may be generated based on at least a portion of the sensor data and / or based on at least a portion of the digital map. For example, a scene may be derived from the surrounding environment model. The scene may also be obtained based on the digital map.

[0054] In a further step S3 , the robustness of the localization system 10 is derived from the scenario ascertained in step S2 and / or the robustness of the localization system is ascertained based on the scenario.

[0055] In a further step S4 , the current robustness of the positioning system 10 is adapted based on the robustness derived from the scenario.

[0056] For this purpose, at least one parameter of the allocation module 18 of the positioning system 10 can be adapted based on the degree of robustness derived from the scenario. This parameter can, for example, describe the number of objects considered in the surroundings of the vehicle. This parameter can also describe the maximum distance to the vehicle up to which the sensor data are evaluated and / or within which the sensor data are considered.

[0057] Furthermore, in step S4, the map data of the digital map and / or the algorithm for assigning the sensor data to the digital map can be selected and / or adapted to match the current robustness. The resolution of the digital map can also be selected and / or adapted. At least one parameter of the algorithm for assigning the sensor data to the digital map can also be selected and / or adapted. The number of particles in the particle filter of the assignment module 18 can also be adapted. Alternatively or additionally, the data processing of the sensor data can be adapted.

[0058] In a further step S5 , at least a portion of the sensor data is assigned to at least a portion of the digital map by means of an assignment module 18 of the localization system 10 based on the robustness of the matching in step S4 .

[0059] In a further step S6 , the position, location and / or orientation of the vehicle is ascertained using position determination module 20 of localization system 10 based on the sensor data assigned to the digital map.

[0060] In an optional step S7, a control command for controlling the vehicle is generated based on the positioning, position and / or orientation determined in step S6. The control command can be output via an interface of the positioning system and, for example, provided to another control device of the vehicle. The control device can then navigate, locate, monitor and / or control the vehicle based on the control command.

[0061] Figure 3 A flow chart for explaining the steps of a method for robustly positioning a vehicle according to an embodiment of the present invention is shown. Figure 3 The method has the same Figure 2 The same steps as the method.

[0062] In step S1 , an initial position determination, ie an initial position and / or initial orientation of the vehicle, is determined, for example, using localization device 16 .

[0063] In step S2 , positioning system 10 is initialized and can be operated, for example, in a standard mode with a predefined degree of robustness, a predefined performance and / or a predefined positioning accuracy of positioning system 10 .

[0064] The position (ie the location and / or orientation) of the vehicle is then determined in step S3 and optionally output in step S3 ′.

[0065] Then in step S4, as in Figure 2 As explained in , the scene can be ascertained based on a digital map and / or based on sensor data from sensors of the vehicle.

[0066] In step S4', as in Figure 2As explained in , the robustness of the positioning system can be derived from the scenario.

[0067] In addition, in step S4 and / or S4', as in Figure 2 As explained in , the current degree of robustness and / or a predefined degree of robustness can be adapted based on an adaptation of at least one parameter of the positioning system 10 .

[0068] The parameters thus matched and / or the degree of robustness of the matching can then be used to localize the vehicle according to step S3 and / or to output the localization in step S3 ′.

[0069] In this case, steps S3 and S4, and optionally S3' and S4', can be executed iteratively. As soon as a scene change is determined in step S4, at least one of steps S4', S3 and S3' can be triggered and / or initiated.

[0070] As a supplement to this, it should be pointed out that "comprising" does not exclude the presence of other elements and "a" does not exclude the presence of a plurality. It should also be pointed out that the features described with reference to any one of the above embodiments can also be used in combination with other features of other embodiments described above.

Claims

1. A method for robustly positioning a vehicle by means of a positioning system (10), the method comprising the following steps: Providing sensor data of at least one sensor of the vehicle, wherein The sensor data contain information about the surroundings and / or the own movement of the vehicle; providing a digital map of the vehicle's surroundings; Characterized in that the method further comprises the following steps: Ascertaining a scene of the vehicle's surroundings based on the sensor data and / or based on the digital map; Derivation of a required robustness of the positioning system (10) from the determined scenario; Adapting a current robustness of the positioning system (10) based on a required robustness of the positioning system derived from the scenario; Based on the robustness of the matching, assigning at least a portion of the sensor data to at least a portion of the digital map by means of an assignment module (18) of the positioning system (10); Based on sensor data assigned to the digital map, a position and / or orientation of the vehicle is ascertained by means of a position determination module (20) of the positioning system (10).

2. The method according to claim 1, wherein: The step of matching the current robustness level comprises matching at least one parameter of an allocation module of the positioning system (10) based on the robustness level derived from the scenario.

3. The method according to claim 2, wherein: The at least one parameter describes the number of objects taken into account in the surroundings of the vehicle; and / or wherein the at least one parameter describes a maximum distance to the vehicle up to which the sensor data are evaluated and / or taken into account.

4. The method according to claim 2 or 3, wherein: The adaptation of at least one parameter of the allocation module (18) comprises at least one of the following partial steps: selecting map data of the digital map to allocate at least a portion of the sensor data; selecting a resolution of the digital map; selecting an algorithm for assigning the sensor data to the digital map; adapting at least one parameter of an algorithm, in particular a map matching algorithm, for assigning the sensor data to the digital map; The number of particles in the particle filter of the distribution module (18) is matched.

5. The method according to claim 1 or 2, wherein: The step of matching the current robustness comprises matching the data processing of the sensor data.

6. The method according to claim 1 or 2, wherein: The ascertained scene of the vehicle's surroundings includes at least one element selected from the group of elements: The obstruction of a portion of the vehicle's surroundings by objects in the surroundings, the number of objects in the vehicle's surroundings, the type of road, intersection, type of intersection, parking lot, residential area, the degree of traffic on the road, the number of other traffic participants in the vehicle's surroundings, the speed of the vehicle, the speed of other traffic participants, and the number of obstacles in the vehicle's surroundings.

7. The method according to claim 1 or 2, wherein: The step of obtaining the scene includes at least one of the following steps: Obtaining a degree of interference and / or a degree of consistency of the sensor data; A validity date of at least a portion of the digital map is determined.

8. The method according to claim 1 or 2, further comprising: generating a model of the vehicle's surroundings based on at least a portion of the sensor data and / or based on at least a portion of the digital map; in, The degree of robustness is derived based on the generated surrounding environment model.

9. The method according to claim 1 or 2, wherein: The digital map has a geo-referenced description of the recommended level of robustness; Wherein the robustness is derived from the geo-referenced description. 10 . The method according to claim 1 , further comprising: ascertaining scene changes and deriving a further degree of robustness based on the ascertained scene changes. 11 . The method according to claim 1 , further comprising: generating a control command for controlling the vehicle based on the ascertained position and / or orientation.

12. A positioning system (10) for robustly positioning a vehicle, wherein: The positioning system (10) is configured to carry out a method according to any one of the preceding claims.

Images