Guidance of a vehicle using environmental data collected from other vehicles

Description

[0001] The invention relates to a method and an arrangement for guiding a vehicle laterally, in particular a motor vehicle, such as a passenger car or truck.

[0002] Modern vehicles, and especially motor vehicles, utilize various driver assistance systems. These often enable autonomous control of vehicle movements and / or autonomous interventions in vehicle operation, such as adjusting speed or steering angle. One example is a lane keeping assist system. In a familiar manner, this system uses sensors to detect the course of the road (or lane) ahead of the vehicle. It then ensures that the vehicle does not unintentionally leave this lane, for example, by autonomously counter-steering when a lane boundary is reached.

[0003] To provide such functions, the vehicle's surroundings are captured by sensors. This process generates so-called environmental data. Environmental data contains evaluable or analyzed information about the environment and, in particular, about predetermined environmental properties. For example, it can specify the course and / or coordinates of lane markings along an upcoming section of road. One way to capture environmental data is through camera sensors, where the desired environmental properties can be determined from the captured camera images using image analysis.

[0004] A typical application of the driver assistance functions described here and / or the autonomous control of a vehicle based on environmental data is lateral vehicle guidance. This can be understood, for example, as keeping the vehicle in a desired position and / or guiding it laterally (generally horizontal and perpendicular to a vehicle's longitudinal axis), particularly autonomously. For instance, a permissible lateral movement corridor can be defined, corresponding to the width of the roadway. Lateral guidance can ensure that this (virtual) movement corridor is not left. If this is imminent, autonomous steering intervention can occur as part of the lateral guidance system. Solutions, and in particular control units and controllers, that enable such lateral guidance are available on the market.

[0005] However, the reliable sensor-based acquisition of environmental data cannot always be guaranteed in certain situations. For example, the sensor may fail, be dirty, or malfunction in other ways. Furthermore, the environmental characteristics to be detected may not always be apparent, for instance, if the road surface is dirty or other environmental conditions exist that cannot be interpreted, or at least not correctly, based on the sensor data.

[0006] From DE 10 2016 215 825 A1 a method for the external provision of map data is known, which is specified in particular with information on driving limit characteristics, wherein the method is carried out for a driving section selected from a large number of driving section segments with the following steps: Identifying static and dynamic objects associated with the selected route segment from received environmental data from at least one reporting vehicle, such that for each identified static and dynamic object a lateral distance to a longitudinal axis of the reporting vehicle capturing the associated environmental data is determined; determining a lane boundary property of the selected route segment depending on the lateral distances of the identified static and dynamic objects in the selected route segment; and providing the map data updated with the lane boundary property externally.

[0007] A similar method of creating a digital map from swarm data of reporting vehicles is known from US 2018 / 0025235 A1.

[0008] There is therefore a need to improve the lateral guidance of vehicles, particularly with regard to the not always reliable sensory perception of the environment. This problem is solved by a method and an arrangement according to the attached independent claims. Advantageous further developments are specified in the dependent claims. It is understood that all of the introductory features and embodiments are also provided for or can apply to the present solution.

[0009] In principle, the solution according to the invention (i.e., the method and arrangement according to the invention) provides that the lateral guidance is not limited to sensor-acquired data. Instead, data from other vehicles that have previously traveled the same route are also taken into account. In particular, so-called swarm data can be used, which contains and stores data from a large number of vehicles over an extended period. This stored data (especially environmental data) serves as a kind of reference, as it indicates how the environment along a current route has already been perceived by other vehicles. In this respect, the vehicle's own sensor data, and especially environmental data, are compared with the stored environmental data.

[0010] According to the invention, the stored data is validated, or in other words, verified, using the vehicle's own acquired sensor data. The already stored environmental data, if validated, is used additionally or alternatively to the environmental data acquired by the vehicle for the (driver-autonomous) lateral guidance of the vehicle. If the validation fails and, for example, the stored environmental data deviates significantly from the sensor-acquired environmental data, this stored environmental data is not considered for the (driver-autonomous) lateral guidance. Instead, the vehicle's own sensor data is used, for example, to react to short-term changes in the environment.

[0011] One advantage is that the multitude of already acquired environmental data, which are expected to have high and / or already verified accuracy, can be used by default for lateral guidance of the vehicle (provided it is plausible). This data can be prioritized over the potentially error-prone and less accurate sensor readings of the vehicle itself. Furthermore, the system can react to short-term changes in the vehicle's surroundings that have not yet been detected by vehicles that have previously traveled the route, as it can then switch to the environmental data acquired by the vehicle currently traveling the route. Such a discrepancy between the stored and acquired environmental data can lead to a failure of the plausibility check, whereupon, according to the invention, the vehicle's own acquired environmental data is selected by default.

[0012] Overall, a solution is provided in which the more accurate or less error-prone environmental data for lateral guidance are automatically selected with an increased probability, thus increasing the operational safety of the vehicle.

[0013] In detail, a method for guiding a vehicle laterally is proposed, comprising the features of claim 1.

[0014] The environmental data may relate in particular to data concerning the road environment and especially an upcoming section of the road. It may specify and / or quantify environmental characteristics as described below. Data acquisition is performed autonomously by the driver using sensors. In particular, a camera, an ultrasonic sensor, a lidar sensor, or similar devices may be used as sensors.

[0015] Any environmental data described herein may include location information. This allows conclusions to be drawn about the location where the corresponding environmental information or data was obtained. The data may be in the form of digital datasets and / or stored in electrical or digital storage devices.

[0016] Environmental data can be output data from sensors used for environmental perception or can be determined from such output data (or output signals). For example, environmental data can be data on predetermined environmental properties, where these properties are determined from image data (as output data) of a camera sensor. This determination, as well as the storage of the acquired environmental data, can generally be performed by a control unit or control system of the vehicle. In principle, all the process steps or procedures described herein can be carried out by a control system of the vehicle, as will be explained in more detail below.

[0017] The stored environmental data can be stored in the vehicle's internal storage system. This could be, for example, the navigation system's memory, or in other words, the memory of a navigation device. Additionally or alternatively, the environmental data can be stored in an external storage device, such as a server (especially a cloud server). The vehicle can then selectively access this server as needed and request the environmental data relevant to a currently driven, planned, or likely route. Communication with the external storage device, and especially the server, can generally be wireless. Specifically, this could be a mobile network connection or an internet connection. At least some of the externally stored environmental data can be temporarily stored in the vehicle.This part may contain environmental data relating to a currently traveled, planned or likely route.

[0018] The stored environmental data may already be at least partially pre-evaluated. For example, it could be an average value for a predetermined environmental property, calculated from all stored environmental data transmitted by individual vehicles. Such an evaluation of expected values ​​and / or averages for predetermined environmental properties, based on all stored, or in other words, collected or accumulated environmental data, can in principle also be performed within the vehicle itself. An environmental property could be the coordinates of predetermined environmental features, such as a center line or lane markings.

[0019] Raw data and / or already evaluated data can therefore be stored as environment data.

[0020] The vehicles that collected and provided the stored environmental data cannot currently travel the route. Rather, it may be a fleet or population of vehicles that have traveled the route in the past. This can occur over an extended period of several days, weeks, or months. This creates a collection of environmental data from a large number of vehicles over a longer period, serving as a kind of knowledge base or database for environmental perception. This differs from solutions where sensor-acquired data is exchanged exclusively between vehicles currently in the same environment and traveling on the same route. Such real-time data exchange requires that another vehicle is actually present in the vicinity.If this is not the case, no improvement in driving safety can be achieved. Furthermore, the problem may arise that this other vehicle suffers from the same environmental problems affecting sensor detection as the vehicle itself, for example, due to heavy snowfall or a dirty road. Therefore, it is not always guaranteed that even if another vehicle is present, the environmental data it captures will be of higher quality than the data captured by the vehicle itself.

[0021] However, it is also possible, for example, to take into account environmental data from other vehicles that are currently also traveling on the same route, in addition to stored environmental data.

[0022] Validating the stored environmental data against the recorded environmental data can involve comparing this data with itself. In particular, at least one environmental property of this environmental data can be compared with itself.

[0023] When we refer to environmental data in the plural, this does not necessarily mean that multiple environmental properties must be recorded. It could also mean that only one environmental property is recorded, but this property is recorded continuously, depending on location and / or along a preceding route. Since this location-dependent environmental property is updated multiple times as the route is traveled, a corresponding number of individual environmental data points are also obtained (e.g., data on an environmental property present at each location or preceding route segment).

[0024] In particular, plausibility checks can identify discrepancies between the stored environmental data and the recorded environmental data. The greater the discrepancy, the less likely the stored environmental data will be considered plausible. Specifically, a maximum permissible deviation threshold can be defined, which must not be exceeded for plausibility checks to be successful. As explained in more detail below, comparisons and / or deviations can be performed based on quantified parameters that are described in, but are also described by, the environmental data, which may include the aforementioned environmental properties.

[0025] The vehicle's lateral control can generally be performed autonomously by the driver. In particular, this can involve the aforementioned guidance within a permissible movement corridor, including any autonomous steering interventions. Known control algorithms can be used to detect impermissible lateral movement of the vehicle based on current movement parameters and, if necessary, to initiate (autonomous) countermeasures. Performing lateral control based on the validated environmental data (or, in the following case, based on the environmental data acquired by the vehicle) can include defining a permissible movement corridor based on the environmental data and / or at least a non-exceedable limit in the lateral direction.For example, lane markings or lane boundaries can be determined from environmental data, and these can define a maximum permissible lateral position for the vehicle. Accordingly, the lateral guidance system can monitor whether the vehicle is in danger of exceeding this limit and then, for example, autonomously counter-steer the vehicle.

[0026] Further training stipulates that the stored environmental data is swarm data. Swarm data can be understood as data collected by a large number of vehicles (e.g., more than 10 or more than 100) over a period of several days (e.g., at least 10, at least 30, or at least 365) or over several trips along the route. These vehicles can operate independently but can still be collectively referred to as a fleet or swarm. The environmental data individually collected by the vehicles in the fleet or swarm can be aggregated and thus collectively referred to as swarm data, and preferably aggregated and / or stored together, for example, in a central database. The common characteristic of the vehicles is that they have traveled the same route, which can be determined using location information from the environmental data.Within the swarm data, environmental data is therefore available for a large number of journeys along the route. Preferably, the vehicles have determined environmental data relating to predetermined environmental properties that are preferably shared. The swarm data can also be stored externally, e.g., on the servers described above.

[0027] According to the invention, it is provided that if the plausibility check fails (for example, because the quality of the recorded environmental data does not meet a required minimum quality), the environmental data recorded by the vehicle is used as environmental data for lateral guidance or, in other words, taken into account.

[0028] It should be noted that, in general, plausibility checks within the scope of this disclosure can be understood as determining and confirming the quality of the environmental data and / or its significance, applicability, accuracy, or general usability. For this purpose, the acquired environmental data can serve as a reference within the scope of this invention.

[0029] If the plausibility check of the stored environmental data fails, this may indicate that, for example, short-term changes have occurred along the route. For safety reasons, the system then switches to the environmental data recorded by the vehicle.

[0030] According to another embodiment, the environmental data relates to different environmental properties, and a separate plausibility check is performed for each property. Consequently, the vehicle's lateral guidance can be based on the validated environmental properties. Conversely, the vehicle can detect environmental properties that have not been validated, allowing for the use of hybrid forms of stored and detected environmental properties for lateral guidance, depending on the plausibility check results. This increases safety and accuracy because each property is individually verified to ensure it accurately reflects the current state.

[0031] In general, it may be stipulated that an environmental property captured with or derived from the environmental data is one of the following: a lane marking or boundary on the left from the perspective of the vehicle (i.e., in the direction of travel); a lane marking or boundary on the right from the perspective of the vehicle; at least one outer edge of the roadway.

[0032] In this context, a roadway can also refer to a single lane of a multi-lane road or route. A roadway marking can accordingly be a lane marking. Generally, a roadway marking can be a center line or a lateral line, applied to the road surface in the usual manner. A roadway boundary can be a physical obstacle that can at least locally prevent a vehicle from leaving the roadway or that can represent a collision barrier for leaving the roadway, e.g., in the form of a bollard, a guidepost, or a guardrail.

[0033] An outer edge of the roadway can correspond to a road edge and / or a transition area to, for example, the unpaved surroundings.

[0034] Further training stipulates that information about the route traveled by other vehicles is also stored. This information is primarily used to verify the plausibility of the stored environmental data against environmental data recorded by the vehicle. For example, environmental data from vehicles that have traveled along the same or a comparable route (or even a section of road) can be considered. It can then be checked whether a route traveled by a vehicle matches the route or road layout ahead, as recorded by the vehicle's own environmental data. If this is not the case, the environmental data from these vehicles can be disregarded, or more precisely, the environmental data recorded by these vehicles cannot be verified.

[0035] The route information can be stored in the form of location coordinates and / or GPS data or other positioning data, or be defined by these.

[0036] According to another embodiment, the plausibility check is performed by determining the quality of the stored environmental data based on the acquired environmental data. Specifically, it can be determined to what extent the environmental data corresponds to and / or deviates from the acquired environmental data. Depending on this, the quality of the stored environmental data can then be determined. The greater the correspondence or the smaller the deviation, the higher the quality can be. In particular, a quantifiable value can be determined as the quality or, in other words, a quality criterion. As explained below, this can be compared with a minimum quality, which can also be defined as a quantifiable value, and the environmental data can only be validated if the minimum quality is met.

[0037] In particular, it can be provided in this context that the quality is determined by comparing the stored environmental data with the acquired environmental data over (i.e., along) a defined and preferably ahead section of the road. Location information for the ahead section of the road can be determined from the currently acquired position of the vehicle and known detection ranges of the vehicle sensors. For example, within the scope of this invention, it is generally possible to determine a considered ahead section of the road, for instance, whether the acquired environmental data is captured within a range of 0 to 20 m, 0 to 40 m, or 0 to 60 m (in the direction of travel) in front of the vehicle.From the stored environmental data, environmental data with identical location information can be determined and compared with the recorded (ahead) environmental data along the upcoming route, and especially within a defined ahead section of the route (e.g., 0 to 20 m or 0 to 40 m). This provides an effective means of detecting impermissible discrepancies between the stored and recorded environmental data in a timely manner.

[0038] The invention can further provide that a minimum quality standard is determined which the stored environmental data must meet in order to be considered or assessed as plausible. This minimum quality standard can be determined depending on at least one of the following (and, more precisely, depending on one of the following parameters): A section of road ahead of the vehicle, within which recorded environmental data is compared with stored environmental data. For example, it can be considered whether the ahead measurement range for environmental data is 0 to 20 m, 0 to 40 m, or 0 to 60 m. The larger the ahead section of road or the considered length of the ahead section, the lower the minimum quality requirements can be. This is based on the idea that measurements in areas far ahead or far from the vehicle are generally less accurate than in areas closer to the vehicle. Therefore, the smaller the aforementioned section size or length, the higher the minimum quality requirements can be.In general, minimum quality can define and, in particular, quantify a minimum level of conformity to be achieved or a permissible maximum deviation from the stored and recorded environmental data. High minimum quality requirements can be synonymous with low permissible deviations or a high required conformity. Low minimum quality requirements can be synonymous with a high permissible deviation and low conformity. A type of environmental property that is recorded and compared as (or within the framework of) environmental data. Consideration can be given to the accuracy with which the corresponding properties can actually be recorded. For example, recording lane boundaries or road markings may be easier than recording a road edge (e.g., the transition from asphalt to natural ground).The higher the theoretically measurable accuracy of an environmental characteristic, the higher the requirements for the minimum quality can be. Conversely, the lower the accuracy, the lower the minimum quality requirements can be. The acquisition conditions during the acquisition of environmental data are crucial, particularly those acquired by the vehicle. These conditions can be determined by vehicle sensors and might include, for example, brightness, precipitation level, visibility, general visibility conditions, or road surface contamination. The acquisition conditions can be stored as part of the environmental data and / or as additional information. If the acquisition conditions are considered critical, i.e.,If highly accurate environmental perception is not possible, as this can be automatically determined and / or evaluated by a vehicle control unit, correspondingly lower minimum quality requirements may apply. Conversely, the minimum quality requirements can be increased if the perception conditions are non-critical. A region within which the vehicle is located. This can be a geographical region defined, for example, by geographical features and / or map data. It can also be a radius with a predetermined number of kilometers. For this region, it can be determined to what extent stored environmental data is available and / or what the (historical) quality of the environmental data in this region has been.If a region is characterized by a small number of stored environmental data points and / or low quality of these data points, the minimum quality requirements can be correspondingly low, as a high degree of uncertainty in data acquisition is generally to be expected. Conversely, if a large number of stored environmental data points and / or high quality are present in this region, the minimum quality requirements can be increased. This includes considering the historical quality of the stored environmental data. Here, previously determined environmental data quality levels can be considered, independent of any connection to specific regions and, for example, in relation to the currently considered vehicle and / or route. If these are comparatively high, the minimum quality requirements can be increased; otherwise, they can be reduced. Finally, it also considers a characteristic of the comparison used to determine the quality.Here, for example, the generally expected accuracy of the comparison method used and / or a (e.g., statistical) accuracy parameter determined in this process can be considered. Generally, the minimum quality requirements can be reduced for less precise or error-prone methods. For example, different (especially lower) minimum quality requirements may apply to a comparison based on least squares than to a general consideration of the correlation between stored and recorded environmental data. With the least squares method, a constant deviation in the data (e.g., a parallel offset between recorded and stored road alignments) is known to lead to large deviations, which can be considered inaccurate.Correlation is less affected by such a constant deviation and therefore preferentially associated with higher minimum quality requirements. Another example of a comparative method is the determination of the Hausdorff distance, where, as with correlation, high minimum quality requirements may apply.

[0039] Depending on the number of variables considered, the minimum quality can be a multidimensional characteristic map or determined using one, with the dimension being determined, for example, by the number of variables. The variables can be weighted or weightable. The weighting can be flexibly adjustable. In this way, a minimum quality can be determined, for example, using a linear combination of potentially weighted individual values, where each individual value is determined as and / or based on a single variable. The individual value can be a scale value obtained as a function of current values ​​of a corresponding variable (e.g., a rating scale of historical quality from 0 (unusable) to 10 (very good)).

[0040] It can also be generally stipulated that stored environmental data is only considered valid for a limited time. If a maximum permissible age is exceeded, this data can be automatically deemed implausible, preferably without considering other (e.g., quality) characteristics.

[0041] The invention also relates to an arrangement for guiding a vehicle laterally, with the features of claim 9.

[0042] The detection device can be a sensor device or a sensor according to any of the variants described herein and generally provided for in a vehicle, with which environmental data can be detected.

[0043] The control unit can be distributed throughout the vehicle and may, for example, comprise or access several individual control components and / or control units. Alternatively, it can be a single, structurally and / or functionally integrated control unit of the vehicle. Access to environmental data can be achieved by accessing a vehicle bus connected to the control unit and the acquisition device. The acquired environmental data can be processed by the control unit and / or the acquisition device, for example, to determine the environmental properties to be acquired as described herein. To access the stored environmental data, the control unit can access an external storage device (e.g., a memory card).via a mobile network connection) or can read this environmental data from a (temporary) storage device in the vehicle, for example, if it is temporarily stored in the vehicle after being read from an external storage device. The control unit can therefore generally be configured to read, request, and / or receive the stored environmental data, in particular from an external storage device. To validate this stored environmental data, a comparison of any type described above can be performed, and / or the quality of this stored environmental data, preferably also a minimum quality standard, can be determined by the control unit.

[0044] To control the vehicle's lateral movement, the control unit can provide known control functions and, in particular, use environmental data to define non-exceeding lateral positions and / or to define a permissible (virtual) movement corridor. Furthermore, the control unit can compare the vehicle's current lateral positions with a corresponding movement corridor or a general movement limit and, if there is an imminent departure from or exceedance of this corridor, initiate driver-autonomous countermeasures, in particular driver-autonomous steering interventions.

[0045] In general, the arrangement may include any further feature and any further component to provide and / or execute all of the functions, operating states, defects, steps, and measures described herein. In general, the arrangement may be configured to execute a procedure of any type described herein. In particular, all further developments and elaborations of the procedure features may also apply to, or be provided for, the identical arrangement features.

[0046] The invention is explained below with reference to the attached schematic figures: Fig. 1 shows a vehicle comprising an arrangement according to an embodiment of the invention, wherein the arrangement performs a method according to an embodiment; and Fig. 2 shows a flow chart of the method, which is executed from the arrangement. Fig. 1 is executed.

[0047] In the following, the same reference symbols can be used across figures for similar or equivalent features.

[0048] In Fig. 1 A vehicle 10 is shown traveling along a road 12 (or route) in the direction of travel F. The vehicle 10 comprises an arrangement 1 according to an embodiment of the invention. The arrangement 1 includes a control unit 14, which in the example shown is configured as a single control unit. The control unit 14 is wirelessly connected via a mobile communication connection to a server external to the vehicle, and more precisely, to a cloud server 16.

[0049] The arrangement 1 also includes a detection device 15, which in the example shown is a camera sensor. This is arranged in the vehicle 10 such that it detects a section of road ahead and outputs image data of the road according to a detection frequency. This image data can be evaluated in the camera sensor itself or by the control unit 14 in such a way as to determine predetermined environmental properties.

[0050] These environmental features are the left lane boundary LF and the right lane boundary RF, viewed from the vehicle's perspective and in the direction of travel F. In the example shown, a roadway is equated with a single lane for vehicle 10. The left lane boundary LF is a median strip, and the right lane boundary RF is a shoulder. The right lane boundary RF can also be marked by a transition area between road 12 and the surrounding area or natural environment.

[0051] The control unit 14 is also configured to determine the current vehicle position, e.g., using GPS data or other position acquisition data. The environmental properties LF and RF can be stored accordingly in a location-based or location-dependent manner. This can be done in a storage device of the control unit 14 (not shown).

[0052] It should be noted that the control unit 14 can comprise at least one processor, in particular a computer processor. This processor can execute program instructions that are stored, for example, in a memory device of the control unit 14 (not shown). By executing these program instructions, the control unit can be caused to perform or initiate all the functions and / or steps and / or measures described herein.

[0053] Within the scope of this disclosure, it is generally possible, and not limited to further details of the exemplary embodiments, to determine and store position information, particularly in the lateral direction, of the ascertained environmental properties and especially of the roadway boundaries LF, RF in an absolute and / or vehicle-external coordinate system. For this purpose, GPS coordinates or other global coordinates can be used, for example. Thus, preferably for each vehicle location, but also in principle independently of this, coordinates of the detected roadway boundaries LF, RF can be determined and stored in an absolute (e.g., global) coordinate system.

[0054] Also shown is a route R that the vehicle will travel in the depicted state. Route R can be determined based on the actual vehicle positions as the vehicle travels along road 12.

[0055] The solution presented assumes that several other vehicles have previously traveled along a comparable route R on road 12 and recorded the same environmental properties LF and RF. It further assumes that these vehicles transmitted the corresponding environmental data to the vehicle-external storage device 16. There, this environmental data can be stored, as explained above, e.g., location-dependently. The actual routes R traveled by the respective vehicles can also be stored.

[0056] If the currently considered vehicle 10 is traveling along road 12, the environmental data currently being recorded by the detection device 16 and control unit 14 can be determined. Likewise, the stored environmental data of those vehicles that have previously traveled along road 12 in this area (i.e., in the same section of the route) can be retrieved from the storage device 16. Relevant environmental data can be determined, for example, based on the routes traveled R and / or a vehicle position, which should correspond as closely as possible to the position and route R of the vehicle 10 currently traveling along road 12.

[0057] The stored environmental data is then compared with the currently acquired environmental data. Specifically, each of the considered environmental properties LF, RF is individually compared with the corresponding stored environmental properties. If it is determined that the stored environmental data, or more precisely, environmental properties, do not correspond to the currently acquired data—which can be determined by the control unit 14 by performing appropriate comparison and evaluation steps—the stored environmental data can be assessed as implausible. For lateral guidance, the currently acquired environmental data can then be used instead, and preferably only this acquired environmental data. Conversely, if plausibility is established, the stored environmental data can be used for lateral guidance, and preferably only this stored environmental data.

[0058] In Fig. 2 This procedure will be explained in more detail using a flowchart. In step S0, which can be performed continuously and is an optional component of the method according to the invention, a multitude of environmental data acquired by vehicles of a (swarm) population or fleet are stored location-dependently in the vehicle-external storage device 16. In step S1, a road 12 is then driven on by a specific vehicle 10. This vehicle 10 also continuously acquires environmental data, and in particular the same environmental properties LF, RF, which have already been acquired by the vehicles whose environmental data were stored in the vehicle-external storage device 16.

[0059] In step S2, the stored environmental data is read or requested by the control unit 14 depending on the current location of the vehicle 10. Subsequently, at least one or any combination of the plausibility checks described below can be carried out.

[0060] Optionally, it can first be determined whether a route R of the vehicles that generated the corresponding stored environmental data matches a preceding route R of the currently viewed vehicle 10. For this purpose, information about route R can be stored as part of the stored environmental data during its acquisition and determined by the control unit 14. A current route R can be determined, for example, using known image analysis algorithms, in particular as the center line between the left and right lane boundaries LF, RF. The comparison of a preceding route R according to stored environmental data (or swarm data) and the acquired preceding route R is performed in Fig. 2This is carried out in step S2a. If sufficient agreement is found, the environmental data can be considered plausible. Agreement can, for example, be expressed as a deviation in the horizontal plane and / or in the spatial coordinates of the corresponding routes R, or be determined based on these.

[0061] Additionally or alternatively, in step S2b, the previously determined environmental properties LF, RF can be compared with the currently recorded environmental properties LF, RF. For this purpose, for example, a coordinate value of the left lane boundary LF, as recorded by the detection device 16, can be recorded over a certain distance. Furthermore, the coordinate value of this lane boundary LF for the same section of the route can be determined from the stored environmental data. In other words, two graphs can be generated in this way, and a correlation and / or distance between these graphs can be evaluated.

[0062] For example, mathematical comparison methods, such as the single-squared error method or a general correlation analysis, can be applied to determine the agreement (or deviation) between the stored and the recorded environmental data, preferably for each individual environmental property. As a result, a quality factor G of the stored environmental data can be determined, where the quality factor is higher the smaller the deviation and / or the greater the agreement with the actually recorded environmental data or environmental properties.

[0063] In step S3, this quality G can then be used to perform the actual plausibility check.

[0064] For this purpose, a minimum quality level M to be met is preferably determined first, which can be carried out at any time and, for example, even before step S2b. The minimum quality level M, as explained above, allows for the determination, depending on various circumstances, of the quality level G that can realistically be expected for the current driving scenario and / or the general operating situation. If high accuracy for acquiring environmental data cannot be expected for the location under consideration or the general operating situation, the minimum quality level M to be met can be chosen accordingly low. Conversely, if the opposite is true, the minimum quality level M to be met can be correspondingly high.

[0065] In step S4, it is then determined whether, in step S3, a comparison of the determined quality and the minimum quality M to be met, or in step S2a, a comparison of the route R, has shown that the stored environmental data is plausible.

[0066] If this is not the case, step S4 specifies that the vehicle's lateral guidance by the control unit 14 should be based on the environmental data acquired by the vehicle 10 itself. If, however, the stored environmental data has been deemed plausible, the lateral guidance is carried out taking this stored environmental data into account, and preferably exclusively based on this stored environmental data and not on the currently acquired data. Reference symbol list

[0067] 1 Arrangement 10 Vehicle 12 Road 14 Control device 15 Detection device 16 Vehicle-external storage device F Direction of travel LF Left lane boundary RF Right lane boundary GG Quality M Minimum quality R Route

Claims

1. A method for laterally controlling a vehicle (10), the method comprising: - capturing environment data of the vehicle (10) by means of detection sensors while driving a route, - receiving stored environment data that was captured, while driving the route, by a plurality of other vehicles (10) that are not currently driving the route; - checking the plausibility of the stored environment data with the aid of the captured environment data; - executing a lateral control of the vehicle (10) based on the plausibility-checked stored environment data, characterized in that if the plausibility check is failed, the environment data captured by the vehicle (10) will be taken into account as environment data for executing the lateral control.

2. The method according to claim 1, characterized in that the stored environment data are swarm data.

3. The method according to either of the preceding claims, characterized in that the environment data relates to different environment properties, and a dedicated plausibility check is carried out for each environment property.

4. The method according to any of the preceding claims, characterized in that at least one of the following environment properties is captured with the environment data: - a left-hand lane marking (LF) or lane boundary, as seen from the vehicle; - a right-hand lane marking (RF) or lane boundary, as seen from the vehicle; - at least one outer lane edge.

5. The method according to any of the preceding claims, characterized in that items of information about a route (R) traveled by the other vehicles are also stored and these items of information are used to check the plausibility of the stored environment data against environment data captured by the vehicle (10).

6. The method according to any of the preceding claims, characterized in that for the plausibility check, a quality (G) of the stored environment data are ascertained with the aid of the captured environment data, in particular wherein the quality (G) is ascertained with the aid of a comparison between the stored environment data and the captured environment data along a defined route section.

7. The method according to claim 6, characterized in that a minimum quality (M) is determined which the stored environment data must meet to pass the plausibility check.

8. The method according to either of claims 6 or 7, characterized in that the minimum quality (M) is determined as a function of at least one of the following: - a route section length ahead of the vehicle (10), within which captured environment data are compared to stored environment data; - a type of environmental property which is captured and compared as environment data; - acquisition conditions during the capture of the environment data; - a region within which the vehicle (10) is located; - a property of the comparison carried out for ascertaining the quality; - a historical quality of the stored environment data.

9. A system (1) for laterally controlling a vehicle (10), comprising: a detection device (15) for capturing environment data of the vehicle (10) and a control device (14) which is configured to: - check the plausibility of the stored environment data captured by a plurality of further vehicles (10) while traveling on a route, with the aid of the captured environment data of the detection device (15); - execute a lateral control of the vehicle (10) based on the plausibility-checked stored environment data, characterized in that the control device is configured to execute a lateral control of the vehicle (10) based on the environment data captured by the detection device (15) if the plausibility check is failed.