Apparatus and method for determining a reference course

Abstract

To efficiently and accurately identify map data for lanes of a digital map that enables reliable and comfortable longitudinal and / or lateral driving of an automated vehicle. The apparatus for identifying a reference course of a core roadway section includes: - to identify the measured travel path 160 of the vehicle 110 for traversing the core roadway portion 410; - assigning a first subset of the measured travel paths 160 to a first subset of the set of distinct subsets, such that the distinct subsets include a core roadway portion 410 and a surrounding roadway portion 410 located immediately before and / or immediately after the core roadway portion 410; and - determining a first reference course 700 of the core roadway portion 410 for the first sub-sequence based on a first subset of the measured travel path 160 for the first sub-sequence; It is composed.

Description

[Technical field]

[0001] The present invention relates to a method and a corresponding device that allows a typical driving path of a vehicle on a roadway to be identified as a driving lane of the driving path. [Background technology]

[0002] To partially or fully automate the control of the vehicle in the longitudinal and / or lateral directions, it is advantageous to have a high-precision geographical map of the surroundings. A map with a simple definition (SD: “standard definition”), which can be used for example for route guidance of the vehicle to a predefined destination, typically has an accuracy in the range of about 1 meter to about 10 meters. A map with a high definition (HD: “high definition”) is typically supposed to differ from reality by less than about 1 meter. Besides the accuracy, HD and SD maps can also differ further, for example with respect to the vertical dimension they each include.

[0003] For example, when a vehicle is traveling on a roadway having multiple lanes (also called driving lanes), the HD map can show lane boundaries to enable a vehicle to distinguish which lane it is traveling in. To identify the lane boundaries, visual indicators such as lane markers (lane markings, lane divisions) can be used. Lane markers usually include lines that are directly provided on the roadway.

[0004] Determining an HD map using a measurement vehicle is time-consuming. Alternatively, an HD map can be created based on sensor detection of lane boundaries, which are determined by a group of vehicles that travel on the roadway anyway.

[0005] Thus, the HD map can display, for a roadway, the course of the lane boundaries between one or more lanes of the roadway. This information can be used by the navigation functions for at least partially or fully automated longitudinal (longitudinal guidance) and / or lateral (lateral guidance) driving of the vehicle along the roadway, in particular along the lanes. For this purpose, a target trajectory for the vehicle can be determined based on the map data of the HD map. The target trajectory can then extend, for example, in the middle between the lane boundaries displayed by the map data. The vehicle can then be automatically longitudinally and / or laterally driven along the determined target trajectory.

[0006] Using lane boundaries to determine the target trajectory may result in the determination of a target trajectory that extends between different lane boundaries but not in the available lanes (e.g., in the case of structurally unsegregated highways). Also, in certain parts of the lane, there may be no lane boundaries detectable by the sensor or there may not be enough of such lane boundaries (e.g., before a toll booth, in an intersection or on a suburban road). As a result, in some cases, a target trajectory or a realistic target trajectory cannot be determined. Also, a target trajectory determined based on the progression of lane boundaries detected by the sensor may be perceived as unpleasant by the user of the automatically driven vehicle. Summary of the Invention [Problem to be solved by the invention]

[0007] The present specification seeks to solve the technical problem of efficiently and accurately identifying map data of digital (HD) maps for lanes that allows reliable and comfortable longitudinal and / or lateral driving of automated vehicles. [Means for solving the problem]

[0008] The problem is solved by the respective independent claims. Advantageous embodiments are described in particular in the dependent claims. It is pointed out that the additional features of the claims dependent on the independent claims can form an independent invention from the combination of all the features of the independent claims, both in themselves and also in combination with the features of the independent claims only with some of the features of the independent claims, which can be made for the subject matter of the independent claims, divisional applications or subsequent applications. This also applies to technical suggestions given in the description that can form an independent invention from one of the features of the independent claims. In particular, the features of the devices and methods described herein can be combined with each other in any way. In particular, the features of the first method or the first device can be applied individually or in combination to another second method or another second device.

[0009] According to one aspect, an apparatus is described for determining a reference driving path for a roadway portion in a road network. The roadway portion may have one or more different lanes. The roadway portion may have a length of, for example, 20 to 50 meters.

[0010] The device is configured to determine a plurality of measured travel paths of one or more vehicles for a corresponding plurality of traverses of the roadway portion, where for each traverse, the measured travel paths can be transmitted from each vehicle to the device via a (wireless) communications connection and received by the device, for example, 5 or more, or 10 or more, or 20 or more measured travel paths can be determined.

[0011] The measured travel path of the vehicle may comprise a series of measurement points of the position of the vehicle, in particular of a predetermined reference point (e.g. of the center of the axis) of the vehicle, respectively, when passing through the roadway section. The measurement points may be provided with a predetermined spatial resolution, for example one measurement point per meter or better. Alternatively or additionally, the measured travel path of the vehicle may indicate the actual travel trajectory of the vehicle when passing through the roadway section. It is thus possible to determine measured travel paths that respectively indicate along which trajectory the individual vehicle travels (manually) through the roadway section. It is thus possible to describe the actual travel behavior of the vehicle on the roadway section by means of the measured travel paths.

[0012] The device is also configured to determine the course of the roadway section into a series of support point planes. In other words, it is possible to arrange the support point planes (for example at equal intervals) along the course of the roadway section. The series of (two-dimensional) support point planes can then be arranged such that directly successive support point planes along the course of the roadway section each have a predefined (constant) distance from one another, for example of 1 to 3 meters. For example, the course of the roadway section can be divided into 10 or more or 20 or more different support point planes. Alternatively or additionally, the series of support point planes can be arranged such that the individual support point planes (at each point of the support point planes) are each arranged perpendicular to the course of the roadway section.

[0013] The device can be configured to determine, for a roadway segment, a map progression of the roadway segment shown in the digital map based on a digital map (e.g. a 2D map). The map progression of the roadway segment can then be used for dividing the roadway segment into a series of support point planes or for arranging the series of support point planes. Alternatively, a reference progression of the roadway segment can be determined (based on a number of measured travel paths) and used for arranging the support point planes, as described below.

[0014] The two-dimensional support plane can be used to jointly analyze a number of measured driving paths to identify at least one reference driving path for the roadway section, which is representative of a typical driving behavior of the vehicle when passing through the roadway section, the reference driving path being capable of describing in particular the centerline of a lane or driving lane of the roadway within the roadway section.

[0015] Furthermore, the apparatus can be configured to identify, for a plurality of measured travel paths, a respective series of intersections of each measured travel path with a respective series of support planes, and thus it is possible to identify where (i.e. at what intersections) each measured travel path passes through each support plane.

[0016] The device may also be configured to identify, for each of the series of support planes, a set of support points (i.e. zero, one or more support points) based on the identified intersection points with each support plane. The set of support points for a support plane may be identified based on a clustering algorithm for clustering the identified intersection points with the support plane, in particular the DBSCAN algorithm. The support points may possibly be identified as the (trimmed) average value of the intersection points of each cluster.

[0017] In some cases, one or more intersections may be identified as outliers when identifying the support points and may remain unconsidered when forming the clusters and / or identifying the support points. In other words, the device may be configured to identify one or more of the intersections with the support plane as outliers. Then, the one or more intersections identified as outliers may remain unconsidered when identifying the set of support points for each support plane.

[0018] The device can therefore be configured to identify a subset (particularly a cluster) of the identified intersections with the support plane for the support point, in particular based on a clustering algorithm, and the support point can then be identified as the average value of the subset of intersections, in particular the trimmed average value, whereby x% of the smallest and / or largest intersections (for each coordinate in the support plane) in the identification of the trimmed average value can be left out of consideration when forming the average value (e.g. x is between 5 and 15).

[0019] The device can also be configured to determine a reference driving path for the roadway section based on a set of support points for a corresponding set of support point planes. For this purpose, the device can be configured to assign, successively along the set of support points, one support point each of the sets of support points of the respective considered support point planes to the reference driving path to be determined.

[0020] For this purpose, the device can be configured to identify, for a first support point in a given support plane of the series of support planes, a first subset of the measured travel paths of one or more intersection points at which the first support point of the given support plane is identified, in particular which intersection points of the one or more measured travel paths are grouped, in particular clustered, to the first support point.

[0021] The device can also be configured to identify a second subset of the one or more measured travel paths that have identified support points of other support planes (already assigned to the reference travel path) (e.g., located immediately before the given support plane). It is therefore possible to identify which of the one or more measured travel paths have been grouped together, in particular clustered, with support points of other support planes that have already been assigned to the reference travel path.

[0022] The first support point can then be assigned to the reference path of travel depending on the first and second subsets of the measured path of travel, in particular depending on a Jaccard coefficient based on the first and second subsets of the measured path of travel. A support point based on a predetermined support point plane can then be assigned to the reference path of travel in which the respective determined subset of the measured path of travel has a large, in particular maximum, overlap with the second subset of the measured path of travel. Correspondingly, it is possible to proceed successively through a series of support point planes.

[0023] Furthermore, the assigned support points of the respective considered support planes can be connected by path segments to the preceding support planes arranged before the considered support plane. Thus, the reference driving paths can be formed successively by the path segments between the respectively assigned support points. Correspondingly, if necessary, it is possible to determine a plurality of reference driving paths for different lanes of the roadway section.

[0024] The reference driving path determined by the device can be formed in such a way that the reference driving path is used as a target trajectory for the at least partially automated driving vehicle in passing through the roadway section. Alternatively or in addition, the determined reference driving path can represent a series of target positions of the vehicle, in particular a series of target positions of a reference point of the vehicle, in passing through the roadway section. Alternatively or in addition, the determined reference driving path can be formed in such a way that the reference driving path can be used in an augmented reality display and / or in estimating the arrival time of the driving route (through the roadway section).

[0025] The device can be configured to determine an empirical speed for a path segment between a first support point in a first support plane and a second support point in a (other) second support plane of the reference driving path. Correspondingly, an empirical speed can be determined for each path segment of the reference driving path. Thus, a typical progression of the (empirical) speed of the vehicle can be determined along the reference driving path. For this purpose, for a plurality of measured driving paths, it is possible to provide speed values ​​along the respective measured driving paths (e.g. speed values ​​for a series of measurement points along the respective measured driving paths).

[0026] To determine an empirical speed for a path segment between a first support point and a second support point, a first speed can be determined for one or more intersections of a first subset of measured travel paths that identify (particularly clustered) the first support point, and a second speed can be determined for one or more intersections of a second subset of measured travel paths that identify (particularly clustered) the second support point, and the empirical speed for the path segment can then be precisely determined based on the first and second speed values, in particular based on an average value of the first and / or second speed values.

[0027] The empirical speed progression along the reference driving path can be taken into account for determining a target trajectory of the at least partially autonomously driving vehicle, thus making it possible to further improve the quality of the autonomous driving function.

[0028] The device can also be configured to provide at least one determined reference driving path as map data for a roadway section for a digital map (in particular for HD maps), whereby for each reference driving path it is possible to provide a progression of the determined experienced speeds along the reference driving path.

[0029] Thus, a device is described that is configured to evaluate a driving path detected (measured) by a vehicle in order to determine one or more reference driving paths that describe the course and absolute position of the (actually driven) driving lane, in particular the course and absolute position of the center line of the driving lane. Here, the device can be configured to determine the reference driving path independently of the roadway markers of the roadway section detected by the sensor. Alternatively or additionally, the device can be configured to determine the reference driving path on the basis of a number of measured driving paths, even if the roadway section does not have roadway markers detectable by the sensor, in particular roadway markers for indicating one or more lanes or driving lanes. Thus, it becomes possible to determine the course of the driving lane on the roadway in a particularly efficient and stable manner.

[0030] According to one aspect, an apparatus is described for determining a reference driving path for a roadway portion (hereinafter a "core roadway portion"), the apparatus being configured to determine (e.g., receive by one or more vehicles) a plurality of measured driving paths of one or more vehicles for a corresponding plurality of traversals of the core roadway portion.

[0031] The device may also be configured to assign each of the measured driving paths to a respective one of the different sets of partial sequences to identify a respective one subset of the measured driving paths for each of the sets of partial sequences. In other words, each measured driving path may be assigned to exactly one partial sequence. Different partial sequences of the set of partial sequences may be identified based on, for example, a digital map (e.g., an SD map) of the road network around the core roadway portion. For example, two or more, or five or more different partial sequences may be considered. Each subset of the measured driving paths may have, for example, two or more, or five or more measured driving paths.

[0032] The different partial sequences may each have a core roadway portion and at least one surrounding roadway portion arranged immediately before and / or immediately after the core roadway portion in the direction of travel. In particular, each partial sequence may comprise, in addition to the core roadway portion, N surrounding roadway portions before the core roadway portion and M surrounding roadway portions after the core roadway portion, where for example N and M are 1 or more or 2 or more, respectively. Here, in some cases N or M may be zero.

[0033] The different partial sequences may comprise at least partially different surrounding roadway portions (i.e. shorter roadway portions) around the core roadway portion. Alternatively or additionally, the different surrounding roadway portions may have different spatial orientations (in particular different driving directions) relative to the core roadway portion.

[0034] Thus, different sub-sequences may have different directions that a vehicle may travel to enter and / or exit the core roadway portion, which may affect the (optimal and / or typical) trajectory of the vehicle on the core roadway portion.

[0035] The apparatus may be configured to determine, for each sub-sequence of the set of sub-sequences, a reference driving path for the core roadway portion based on (and specifically limited to) a respective subset of the measured driving paths. In other words, the determination of the reference driving path in the core roadway portion for a given sub-sequence may be limited to a determined subset of the measured driving paths for the given sub-sequence.

[0036] To identify the reference driving path, the methods described herein can be used. In particular, the apparatus can be configured to identify the reference driving path for the core roadway portion such that the reference driving path is defined and / or described by a series of support points in a corresponding series of support point planes disposed along the course of the core roadway portion.

[0037] The apparatus may be configured to identify, for each measured travel path of a subset of the measured travel paths for the predetermined sub-sequence, a respective series of intersections of the respective measured travel paths with a corresponding series of support point planes, to identify a reference travel path for the predetermined sub-sequence. Then, for each of the series of support point planes, a respective set of support points may be identified based on the identified intersections with each support point plane. Also, for the predetermined sub-sequence, a reference travel path in the core roadway portion may be identified based on the set of support points for the corresponding sequence of support point planes.

[0038] Thus, a device is described that allows different entry and / or exit directions of a core roadway section to be taken into account when evaluating the measured driving path. For each of the different entry and / or exit directions, a dedicated reference driving path can then be determined. Thus, the accuracy of the determined course of the driving lane can be efficiently improved.

[0039] The device can be configured to arrange the series of support point planes such that the first support point plane of the series (the first support point plane in the direction of travel through the core roadway section) has a predefined distance (for example, half the distance between the support point planes) with respect to the beginning of the course of the core roadway section. Alternatively or in addition, the device can be configured to arrange the series of support point planes such that the last support point plane of the series (the last support point plane in the direction of travel through the core roadway section) has a predefined distance (for example, half the distance between the support point planes) with respect to the end of the course of the core roadway section. Thus, it is possible to efficiently and stably improve the quality of the link of the reference travel path for consecutive roadway sections.

[0040] The device can be configured to provide a set of the determined reference driving paths for the corresponding sets of different partial sequences as map data for the core roadway portion for a digital map (e.g., for an HD map), thus further improving the quality of the automated driving function (based on the provided map data).

[0041] The device may be configured to identify, for each of the plurality of measured travel paths, a respective extended measured travel path that describes the passage of the core roadway portion as well as the passage of at least one surrounding roadway portion located immediately before and / or after the core roadway portion. In particular, an extended measured travel path may be identified that extends over the length of the sub-sequence. The plurality of measured travel paths may then be efficiently and precisely assigned to individual sub-sequences of the set of different sub-sequences based on the corresponding plurality of extended measured travel paths.

[0042] The device can be configured to determine a subsequent reference driving path for a specific peripheral roadway portion for a specific partial sequence having a core roadway portion and a specific peripheral roadway portion arranged immediately after the core roadway portion. The reference driving path for the core roadway portion can then be combined with the subsequent reference driving path for the specific peripheral roadway portion to determine a sequence driving path for the specific partial sequence. Alternatively, a chain of reference driving paths in the reverse direction can be performed, so that the reference driving path for the core roadway portion is combined with a previous reference driving path for a specific peripheral roadway portion arranged immediately before the core roadway portion. Thus, a continuous driving path through the road network and thus the course of the driving lanes can be determined efficiently and precisely.

[0043] The device can be configured to determine a set of subsequent reference driving paths for a predetermined surrounding roadway portion located immediately after the core roadway portion. The device can also be configured to assign the subsequent reference driving paths of the set of reference driving paths for the predetermined surrounding roadway portion to the predetermined reference driving paths of the set of reference driving paths for the core roadway portion using an (assigned) clearance. In this case, the (assigned) clearance can depend on the distance between the end point (i.e., the position of the end point) of the predetermined reference driving path and the start point (i.e., the position of the start point) of the subsequent reference driving path. The predetermined reference driving path can then be combined (by a path segment) with the assigned subsequent reference driving path to determine a sequence driving path for a series of roadway portions. Thus, it is possible to efficiently and accurately determine a continuous driving path through a road network, and thus the course of a driving lane.

[0044] The clearance may depend in particular on the Euclidean distance between the end point of the predetermined reference travel path and the start point of the subsequent reference travel path. Alternatively or in addition, the clearance may depend on the Jaccard coefficients of a first subset of measured travel paths and a second subset of measured travel paths. In this case, the first subset of measured travel paths may include one or more measured travel paths, the corresponding one or more intersection points of each series of support point planes with the last support point plane are grouped, in particular clustered, to the end point of the predetermined reference travel path. The second subset of measured travel paths may include one or more measured travel paths, the corresponding one or more intersection points of each series of support point planes with the first (initial) support point plane are grouped, in particular clustered, to the start point of the subsequent reference travel path.

[0045] In other words, the device can be configured to identify a first subset of the measured travel paths including one or more measured travel paths (only) that identify an end point of the predetermined reference travel path. The device can also be configured to identify a second subset of the measured travel paths including one or more measured travel paths (only) that identify a start point of the subsequent reference travel path. Then, a clearance for a distance between an end point of the predetermined reference travel path and a start point of the subsequent reference travel path can be accurately identified based on the first and second subsets of the measured travel paths, in particular based on a Jaccard coefficient based on the first and second subsets of the measured travel paths.

[0046] According to another aspect, a control unit for at least partially automated longitudinal and / or lateral driving of a (motorized) vehicle along a driving route through a road network is described. The control unit may be part of the vehicle. The device may be configured to determine (e.g. based on map data of a digital map of the road network) a set of multiple different reference driving paths for a corresponding set of different partial sequences for a preceding passage of a core roadway portion of the road network in the driving route, where the different partial sequences may each comprise a core roadway portion and at least one surrounding roadway portion located immediately before and / or immediately after the core roadway portion in the driving direction of the vehicle.

[0047] The control unit is also configured to identify a partial sequence from the set of different partial sequences based on the travel route, whereby it is possible to identify in particular a partial sequence that corresponds to the travel route of the vehicle, i.e. along which the vehicle travels on the core roadway portion.

[0048] Furthermore, the control unit can be configured to determine a target trajectory of the vehicle for at least partially automated longitudinal and / or lateral driving during passage through the core roadway portion based on the reference driving path for the identified partial sequence. The vehicle can then be driven longitudinally and / or laterally in an at least partially automated manner along the determined target trajectory through the core roadway portion. Thus, the quality of the automated driving function can be efficiently and reliably improved.

[0049] According to another aspect, an apparatus for determining a reference driving path for a roadway portion is described. The apparatus can be configured to determine (e.g., receive by one or more vehicles) a plurality of measured driving paths of one or more vehicles for a corresponding plurality of traversals of the roadway portion.

[0050] Additionally, the apparatus may be configured to determine a reference course of the roadway portion based on a plurality of measured travel paths.

[0051] To this end, the device can be configured to determine a respective total length based on at least a subset of the plurality of measured travel paths, the subset may include, for example, 70% or more of the measured travel paths of the plurality of measured travel paths, and the subset may include, for example, 2 or more, or 5 or more measured travel paths.

[0052] Then, for each measured travel path of the subset of measured travel paths, a corresponding series of consecutive extended positions can be identified, where the consecutive extended positions may be relative to the total length of the measured travel path (e.g. at a series of percentage points of the respective total length). For example, the extended positions may correspond to y% of the respective total length. And the series of consecutive extended positions may have different (possibly equally spaced) values ​​of y (e.g. 10, 20, 30,...100). The series of consecutive extended positions may include, for example, 5 or more, or 10 or more different extended positions.

[0053] In order to identify a series of path points of the reference course at a corresponding series of consecutive extended positions of the reference course, the device can be configured to identify, for each extended position, a corresponding path point of the reference course based on each path point of the measured travel path, in particular based on an average value, e.g. a trimmed average value, of the path points of the measured travel path. The reference course can then be described by the series of path points.

[0054] The device can be configured to identify one or more path points of the measured travel path as outliers for one (each individual) extension position of the series of consecutive extension positions. The one or more identified path points of the measured travel path can remain unconsidered when determining the corresponding path points of the reference course. Thus, the quality of the determined reference course can be further improved.

[0055] The device can be configured to identify one or more of the measured travel paths as outliers. One or more identified path points of the measured travel path can remain unconsidered when determining the reference course of the roadway section. Thus, the quality of the determined reference course can be further improved.

[0056] The apparatus can also be configured to position a series of support planes along the reference path, and for the plurality of measured travel paths, identify a respective series of intersections of each measured travel path with a corresponding series of support planes, and identify a reference travel path for the roadway section based on the plurality of identified series of intersections for the corresponding plurality of measured travel paths.

[0057] By determining the reference course for the roadway section, it is possible to further improve the quality of the determined reference driving path, so that in particular it is possible to determine reference driving paths for successive roadway sections which can be successively linked together to determine a sequential driving path.

[0058] The device can be configured to determine, for a roadway segment, a map progression of the roadway segment shown in the digital map based on a digital map (e.g., an SD map), and a reference progression of the roadway segment can also be determined based on the map progression with increased accuracy.

[0059] The device can in particular be configured to identify the start and / or end of the map course (i.e. the position of the start and / or end) and, based on the clearance, the corresponding start and / or the corresponding end of the reference course (i.e. the position of the start or end, respectively). The start and / or the end of the reference course can then be identified in such a way that the clearance is reduced, in particular minimized. Thus, the quality of the identified reference course can be further improved.

[0060] The clearance for locating the beginning of the reference course may depend on the distance between the beginning of the map course and the beginning of the reference course to be located, and may also depend on the deviation of the orientation of the map course at the beginning of the map course and the orientation of the reference course at the beginning of the reference course to be located.

[0061] Correspondingly, the clearance for locating the end of the reference run may depend on the distance between the end of the map run and the end of the reference run to be located, and may also depend on the deviation of the orientation of the map run at the end of the map run and the orientation of the reference run at the end of the reference run to be located.

[0062] The use of such clearances makes it possible to specify reference travel paths which can be linked particularly reliably and consistently.

[0063] According to another aspect, an apparatus for determining a reference path for a core roadway portion is described. The apparatus can be configured to determine a specific reference path for each of the core roadway portions for one or more different sub-sequences (each of which includes the core roadway portion). The specific reference path for a given sub-sequence can then be used to determine a reference driving path for the given sub-sequence. Thus, the quality of the determined reference driving path for the given roadway portion can be further improved.

[0064] Alternatively or supplementarily, the device may be configured to provide one or more determined reference paths for the core roadway portion as map data of the digital map (as an alternative or supplement to the map path). Alternatively or supplementarily, the device may be configured to operate the vehicle in dependence on the determined reference path of the core roadway portion when traveling through the core roadway portion.

[0065] The apparatus can be configured to identify a plurality of measured travel paths of the one or more vehicles for a corresponding plurality of traversals of the core roadway portion. Further, the apparatus can be configured to assign a first subset of the measured travel paths of the plurality of measured travel paths to a first subset of the set of distinct subset sequences, where each distinct subset sequence can include the core roadway portion and at least one surrounding roadway portion located immediately before and / or immediately after the core roadway portion in the direction of travel.

[0066] Then, based on a first subset of the measured travel paths for the sub-sequence (possibly only the first subset), a first reference path of the core roadway portion for the first sub-sequence can be determined, and correspondingly, specific reference courses for one or more other sub-sequences can be determined.

[0067] The first reference course for the first partial sequence can be determined as described herein, whereby a first subset of the measured travel path (possibly only the first subset) is taken into account for the determination of the first reference course. Thus, the first reference course can be specifically adapted to the course of the first partial sequence. As a result, the quality of the determined first reference course and the quality of the reference travel path determined on the basis of the first reference course can be improved (especially with regard to the chain of reference travel paths for the partial sequence).

[0068] It is thus possible to determine for different partial sequences a respective specific reference course of the core roadway section, which in turn can be used to determine the corresponding partial sequence specific reference driving path, thus making it possible to significantly improve the quality of the determined reference driving path for the core roadway section.

[0069] The first partial sequence may, for example, comprise a core roadway portion and a predetermined surrounding roadway portion located immediately after and / or immediately before the core roadway portion. The device may be configured to determine a reference driving path for the core roadway portion based on a first subset of the measured driving paths and by using a first reference course (as described herein). The device may also be configured to determine a subsequent or preceding reference driving path for the predetermined surrounding roadway portion. The reference driving path for the core roadway portion may then be combined with the subsequent or preceding reference driving path for the predetermined surrounding roadway portion to determine the sequence driving path for the first partial sequence. Using the first reference course to determine the reference driving path for the core roadway portion allows a particularly accurate determination of the sequence driving path.

[0070] The device can be configured to determine an empirical speed for a segment of a first reference course between a first path point and a second path point (immediately following) the first reference course, where the first path point can be located at a first extension position and the second path point can be located at a second extension position. Correspondingly, for every pair of path points (directly succeeding each other), an empirical speed (i.e. an empirical speed value) can be determined. Thus, it is possible to determine an empirical speed course of the vehicle along the reference course of the roadway section.

[0071] The empirical speed between the first and second path points can be determined as follows: a first speed value can be determined for one or more corresponding first path points of a first subset of the measured travel path that identified the first path point of the first reference course; a second speed value can be determined for one or more corresponding second path points of the first subset of the measured travel path that identified the second path point of the first reference course; and an empirical speed for a segment of the first reference course can be determined based on the first and second speed values, in particular based on an average value (possibly trimmed average value) of the first and / or second speed values.

[0072] The device can be configured to provide the first reference course, in particular together with an empirical speed course along the first reference course, as map data for the core roadway portion for the digital (HD) map, so that automated driving functions can be efficiently performed based on the digital (HD) map.

[0073] The reference course provided for a roadway section may refer to the course of the roadway section as a whole (without division into different lanes or driving lanes) resulting from the measured driving path for this roadway section. It is therefore possible to efficiently determine and provide a (relatively rough) course of the roadway section (possibly depending on the preceding and / or following driving direction of the vehicle, i.e. possibly depending on the respective section sequence).

[0074] On the other hand, the reference driving paths determined and provided for the roadway portion may each relate to different lanes and / or driving lanes of the roadway portion, so that by means of the reference driving paths for the roadway portion it is possible to precisely describe different (effective) lanes and / or driving lanes of the roadway portion (possibly depending on the preceding and / or following driving direction of the vehicle, i.e. possibly depending on the respective part sequence).

[0075] Thus, typically only one reference course is determined for a roadway section (and possibly for a given partial sequence) (even if the roadway section has multiple lanes and / or driving lanes), whereas multiple reference courses are determined for a roadway section (and possibly for a given partial sequence), possibly for different lanes and / or driving lanes.

[0076] According to another aspect, a (road vehicle) motor vehicle (in particular a car or a lorry or a bus or a motorcycle) is described that includes one or more of the devices and / or control units described herein.

[0077] According to another aspect, a central unit, e.g. a server, is described that includes one or more of the devices described herein.

[0078] According to another aspect, a method for determining a reference driving path for a roadway section is described. The method includes determining a plurality of measured driving paths of one or more vehicles for a corresponding plurality of passages of the roadway section and dividing the course of the roadway section into a series of support point planes. The method also includes determining, for the plurality of measured driving paths, a respective series of intersections of the respective measured driving paths with the respective corresponding series of support point planes. The method further includes detecting, for each of the series of support point planes, a respective set of support points based on the determined intersections with each support point plane, and determining at least one machine gun driving path for the lane section based on the set of support points for the corresponding series of support point planes.

[0079] According to another aspect, a method for determining a reference driving path for a core roadway portion is described. The method includes determining a plurality of measured driving paths of one or more vehicles for a corresponding plurality of passages of the core roadway portion. The method also includes assigning each of the measured driving paths to a respective one of the set of different partial sequences to determine, for each partial sequence of the set of partial sequences, a respective one subset of the measured driving paths. In this case, each of the different partial sequences may have the core roadway portion and at least one surrounding roadway portion located immediately before and / or immediately after the core roadway portion in the driving direction. Additionally, the method includes determining, for each partial sequence of the set of partial sequences, a reference driving path for the core roadway portion based on a respective subset of the measured driving paths.

[0080] According to another aspect, a method for at least partially automated longitudinal and / or lateral driving of a vehicle along a driving route through a road network is described. The method includes determining, for a preceding passage of a core roadway portion of the road network in the driving route, a set of different reference driving paths for a corresponding set of different partial sequences, in particular based on map data of a digital map of the road network, where the different partial sequences may each comprise the core roadway portion and at least one surrounding roadway portion located immediately before and / or immediately after the core roadway portion in the driving direction of the vehicle. Furthermore, the method includes identifying a partial sequence of the set of different partial sequences based on the driving route, and determining a target trajectory of the vehicle for at least partially automated longitudinal and / or lateral driving during the passage of the core roadway portion based on the reference driving paths for the identified partial sequences.

[0081] According to another aspect, a method for determining a reference travel path for a roadway section is described. The method includes determining a plurality of measured travel paths of one or more vehicles for a corresponding plurality of passages of the roadway section, and determining a reference course of the roadway section based on the plurality of measured travel paths. The method also includes arranging a series of support point planes along the reference course, and determining, for each of the plurality of measured travel paths, a series of intersection points of the respective measured travel path with the corresponding series of support point planes. Additionally, the method includes determining at least one roadway section for the roadway section based on the plurality of determined series of intersection points for the corresponding plurality of measured travel paths.

[0082] According to another aspect, a method for determining a reference course for a core roadway portion is described. The method includes determining a plurality of measured travel paths of one or more vehicles for a corresponding plurality of passages of the core roadway portion. In addition, the method includes assigning a first subset of the measured travel paths of the plurality of measured travel paths to a first sub-sequence of a set of different sub-sequences, where each of the different sub-sequences can include the core roadway portion and at least one surrounding roadway portion located immediately before and / or immediately after the core roadway portion in the travel direction. The method further includes determining a first reference course of the core roadway portion for the first sub-sequence based on the first subset of the measured travel paths for the first sub-sequence.

[0083] According to another aspect, a software (SW) program is described, the software program being configurable to be executed in a processor and thereby to perform one or more of the methods described herein.

[0084] According to another aspect, a storage medium is described. The storage medium can include a software program configured to execute in a processor and thereby perform one or more of the methods described herein.

[0085] It should be noted that the methods, devices and systems described herein can be used alone or in combination with other methods, devices and systems described herein. Furthermore, aspects of the methods, devices and systems described herein can be combined with each other in various ways. In particular, the features of the claims can be combined in various ways. Also, features shown in parentheses should be understood to be optional features.

[0086] The present invention will be described in detail below with reference to examples. [Brief description of the drawings]

[0087] [Figure 1a] FIG. 1 illustrates an example system for determining a reference driving path for a roadway segment. [Figure 1b] FIG. 1 illustrates an example roadway portion having multiple efficient lanes. [Figure 2a] FIG. 1c shows an exemplary measured travel path for the roadway portion shown in FIG. [Figure 2b] FIG. 2 illustrates exemplary intersection and support points in one plane of a roadway section. [Figure 2c] FIG. 2 illustrates exemplary support points for a series of planes of a roadway section. [Figure 2d] FIG. 1c shows an exemplary reference driving path for the roadway portion shown in FIG. [Diagram 3] FIG. 2 illustrates a flowchart of an exemplary method for determining a reference driving path for a roadway segment. [Figure 4a] FIG. 2 illustrates an exemplary measured driving path at a roadway node. [Figure 4b] FIG. 2 is a diagram showing an example chain of reference driving paths for successive roadway portions; [Figure 5a] FIG. 2 illustrates a flowchart of an exemplary method for determining a reference driving path for a roadway segment. [Figure 5b]FIG. 2 illustrates a flowchart of an exemplary method for determining a target trajectory for a vehicle. [Figure 6] FIG. 13 shows exemplary support points for a reference driving path on successive roadway sections. [Figure 7] FIG. 2 illustrates an exemplary series of planes along a reference course of a roadway section. [Figure 8a] FIG. 2 illustrates a flowchart of an exemplary method for determining a reference driving path for a roadway segment. [Figure 8b] FIG. 2 shows a flowchart of an exemplary method for determining a reference course of a roadway section. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0088] As explained at the beginning, this document deals with the efficient and reliable determination of one or more reference driving paths for a roadway portion of a road network. For example, different reference driving paths can be determined for different lanes of the roadway portion. The determined reference driving paths can be provided as map data in an HD map for the road network. In particular, the determined reference driving paths can be used for automated longitudinal steering (longitudinal guidance) and / or lateral steering (lateral guidance) of the vehicle to determine a target trajectory.

[0089] 1a shows an exemplary system 100 for determining at least one reference driving path for at least one roadway portion of a road network. The system 100 includes a central unit 120 configured to determine measurement data 115 for a plurality of vehicles 110 and / or for a plurality of passages of the roadway portion. The measurement data 115 for the passages and / or of the vehicles 110 can be received from each vehicle 110 via a (wireless) communication connection 101. The measurement data 115 of the passages of the vehicles 110 can represent a measured driving path of the vehicles 110 through the roadway portion (e.g. as a series of measurement points (sequence of measurement points) along the driving path traveled by the vehicle).

[0090] The central unit 120 (also generally referred to herein as device) can be configured to determine at least one reference driving path for the roadway portion based on the measurement data 115 of the multiple vehicles 110 and / or multiple passes of the vehicle portion. The reference driving path can represent a typical driving path of the vehicle 110 when passing through the roadway portion. Different reference driving paths can then be determined for different lanes of the roadway portion. The one or more determined reference driving paths for the roadway portion can be provided by the central unit 120 as map data 125 of a digital (HD) map for the roadway portion (e.g., can be transmitted to the one or more vehicles 110 via the communication connection 101).

[0091] Vehicle 110 may include a position sensor 114 configured to detect data about the position of vehicle 110 based on a Global Navigation Satellite System (GNSS). Vehicle 110 also includes one or more sensors 112 (such as, for example, a speed sensor, an inertial measurement unit (IMU), one or more cameras, wheel speed sensors, steering sensors, etc.) configured to detect sensor data that may be used for odometry-based determination of the position of vehicle 110.

[0092] The control unit 111 of the vehicle 110 can be configured to determine the position of the vehicle 110 when passing through the roadway section based on data from the position sensor 114 and / or based on sensor data from the one or more sensors 112. In particular, a series of measurement points of the position of the vehicle 110 can be determined along the course of the roadway section, the series of measurement points representing a travel path actually traveled by the vehicle 110. In this specification, the travel path is referred to as a "measured travel path". The series of measurement points of the measured travel path can be provided by the vehicle 110 (e.g. via the communication unit 113 of the vehicle 110) as measurement data 115. The series of measurement points can for example have a sampling rate or spatial resolution of one measurement point per meter or more.

[0093] 1b shows an exemplary measured travel path 160 for a roadway 150 (which may be divided into one or more roadway portions). Each individual measured travel path 160 includes a series of measurement points 161. The roadway 150 may have pavement markings 152 that indicate optically and / or sensor detectably the edges of the roadway and / or the division of the roadway 150 into multiple lanes 151.

[0094] The driving path 160 on which the vehicle 110 actually travels may correspond approximately to the centerline of the lane 151 of the roadway 150, especially if the roadway 150 has a straight course. On the other hand, for example before a curve, there may be a significant deviation between the centerline of the lane 151 and the driving path 160 on which the vehicle 110 actually travels. Furthermore, a certain roadway 150 does not have road markings for division into multiple lanes 151, for example before a toll booth or at an intersection. In this case, it is not possible to determine the centerline of the lane 151 as the reference driving path based on the road markings 152 detected by the sensor.

[0095] As explained at the beginning, the centerline of the lane 151 can be used as a target trajectory for automated longitudinal and / or lateral driving of the vehicle 110. As illustrated in Fig. 1b, it may happen that in certain roadway sections, there are no roadway markings 152 for identifying the lane 151, and therefore the target trajectory cannot be identified. Also, in certain roadway sections (e.g. before and after a curve), the centerline of the lane 151 may be unsuitable as a target trajectory in order to generate a driving behavior that is typical and appropriately comfortable for a manually driven vehicle 110 during automated driving.

[0096] In Fig. 2a, a number of travel paths 160 determined and measured for a roadway section according to Fig. 1b are shown. Here, it is assumed that the roadway 150 illustrated in Fig. 1b corresponds to the (only) roadway section. The central unit 120 can be configured to determine a course 270 of the roadway section based on a digital (SD) map. Said course 270 can be called a map course of the roadway section. The map course 270 can be used as a reference for the division of the roadway section into a series of (support point) planes (levels) 271. In particular, the planes 271 can be arranged at a given distance (interval) (for example 1 to 3 meters, for example 2 meters) perpendicular to the map course 270, respectively, which are used to divide the measured travel path 160 into a series of intersections in an appropriate manner, respectively.

[0097] Each plane 271 intersects a different measured traveled path 160. Thus, the measured traveled path 160 can be divided by the series of planes 271 into a corresponding series of intersections along the map course 270 of the roadway portion.

[0098] 2b shows a number of intersections 272 for the corresponding measured driving paths 160 in an exemplary plane 271 of a series of planes 271 for the roadway section. The central unit 120 can be configured to convert the number of intersections 272 of the plane 271 into a set of support points 200 for a corresponding number of reference driving paths for the roadway section using a clustering algorithm. As the clustering algorithm, for example, the DBSCAN algorithm can be used.

[0099] The central unit 120 can be configured to identify in an appropriate manner a number of intersections 272 and a set of support points 200 based thereon for a series of planes 271 of the roadway portion, as shown in Fig. 2c. In this case, the number of support points 200 in the different planes 271 can at least partially differ from each other. A change in the number of support points 200 in the different planes 271 can be due, for example, to travel paths merging with each other along the roadway portion or a travel path splitting into multiple travel paths along the roadway portion (which can be the case, for example, for travel to a toll booth).

[0100] As can be seen from Fig. 2c, based on the measured travel paths 160 for the individual planes 271 of the series of planes 271, a set of one or more support points 200 can be determined. Based on the set of the series of support points 200, the central unit 120 can be configured to determine one or more reference travel paths 210 for the roadway section (as exemplarily illustrated in Fig. 2d). For this purpose, it is possible to connect one support point 200 of each plane 271 successively from a first plane 271 along the series of planes 271 (in the travel direction of the roadway section) through a path segment 211 to one support point 200 of the plane 271 directly following each. Thus, in order to determine the reference travel path 210, at least a series of support points 20 for the corresponding series of planes 271 can be determined and can be connected successively to the path segment 211. In the example illustrated in Fig. 2d, a total of five different reference travel paths 210 can be determined, which overlap in partial segments of the roadway section.

[0101] As mentioned above, the centerline of a travel lane (or lane) 151 typically cannot be detected directly by a sensor. As such, the centerline of a travel lane 151 is typically defined and identified as the center between adjacent travel lane boundaries and / or travel lane markings 152. However, this may lead to identifying a travel lane centerline between markings 152 that define a travel lane 151 that does not actually exist at all. This may be the case, for example, for the centerline between inbound and outbound traffic on a highway that is not structurally separated. The measures described herein may reliably avoid identifying non-existent travel lane centerlines, thereby improving the quality of autonomous driving functions.

[0102] Within sizable highway toll booths, the inner reaches of intersections, or on many suburban roads, there are no driving lane markings 152 around the perimeter required to identify the driving lane centerline based on the driving lane markings. The same is true for separate driving lanes 151 that are not separated by markings 152 for each turn. The measures described herein allow for accurate identification of the reference driving path 210 for roadways 150 that do not have lane markings 152 or that have poorly defined lane markings 152.

[0103] Human drivers typically do not follow the driving lane centerline of the driving lane 151 in all circumstances and optimize the trajectory of the vehicle 110 from perspectives other than the clearance to adjacent traffic in one or more adjacent driving lanes 151. Examples of this are shortcuts, maneuvering around turns, or avoiding potholes and other irregularities. Therefore, the driving lane centerline is often not optimal as a target quantity (i.e., a target trajectory) for trajectory planning of the vehicle 110. The measures described herein allow the identification of a reference driving path 210 that can be used for optimized trajectory planning (because the reference driving path 210 represents a typical driving behavior of manual driving).

[0104] The discrete measurement points 161 along the measured trajectory 161 (i.e., the measured driving path) of the actual passage of the roadway portion are processed by the method described herein to identify, for each driving lane 151, a reference driving path 210 that can replace the driving lane centerline as a target quantity for trajectory planning as a representation of natural driving behavior, regardless of the presence of driving lane markings 152.

[0105] For each pass there is one sequence 160 of estimated and / or measured vehicle positions 161 (i.e., measured traveled paths 160). A clustering algorithm can be used to assign intersections 272 of the measured traveled paths 160 with a plane 271 perpendicular to the road course 270 to corresponding intersections 272 of other measured traveled paths 160.

[0106] The individual passages of the roadway or road sections typically differ due to natural driving behavior and due to a number of environmental influences. In order to improve the reliability and accuracy of the determination of the reference driving path 210 along the driving lane, measures can be taken to reduce the influence of disturbances. In particular, when the intersections 272 of the measured trajectory 160 with a plane 271 perpendicular to the road course 270 (possibly for each coordinate) are collected, measures can be taken to form an average value (in particular a so-called trimmed mean) excluding outliers. The central unit 120 can thus recognize outliers when determining the support points 200 and not take them into account in order to improve the accuracy of the determined support points 200.

[0107] 3 shows a flow chart of a (possibly computer-implemented) method 300 for determining a reference driving path 210 for a roadway portion 410 (of a road network). The roadway portion 410 is, for example, shown in FIG. 4a. The method 300 comprises determining 301 a number of measured driving paths 160 of one or more vehicles 110 for a number of corresponding traversals of the roadway portion 410. Each of the measured driving paths 160 may then represent a trajectory that an individual vehicle 110 actually travels during a respective traversal of the roadway portion.

[0108] The method 300 further comprises dividing 302 the path 270,700 of the roadway portion 410 (e.g. the map path 270 or the reference path 700 described in relation to FIG. 7) into a series of support point planes 271. In other words, along the path 270,700, the support point planes 271 (each perpendicular to the path 270,700) can be arranged (at equal intervals). In this case, each of the support point planes 271 can be two-dimensional. The path 270,700 of the roadway portion 270 and / or the measured travel path 160 can be described in a three-dimensional space.

[0109] Additionally, the method 300 includes, for each of the plurality of measured travel paths 160, identifying 303 a series of intersections 272 of each measured travel path 160 with a corresponding series of support planes 271. In other words, it is possible to identify where each individual measured travel path 160 intersects with the support plane 271, and thus forms an intersection 272.

[0110] The method 300 further comprises, for each of the series of support planes 271, identifying 304 a set of support points 200 based on the identified intersection points with each support plane 271. To this end, in each respective support plane 271, it is possible to identify one or more clusters of intersection points 272 based on a clustering algorithm, and based on each cluster, it is possible to identify a respective support point 200 (e.g. as the average (possibly trimmed mean) value of the intersection points 272 in the cluster).

[0111] Furthermore, the method 300 includes determining 305 at least one reference driving path 210 for the roadway portion 410 based on a collection of the series of support points 200 for the corresponding series of support point planes 271. For this purpose, one or more related series of support points 200 can be determined along the series of support point planes 271 and can be connected to each other via respective path segments 211. Each related series of support points 200 then forms one respective reference driving path 210.

[0112] At the branching of the roadway 150 into two different roadways connecting with different directions (e.g. at an exit or intersection), the measured travel path 160 may be different for the vehicle 110 at the roadway portion, depending on whether the vehicle 110 travels into the first roadway connecting after the roadway portion or into the second roadway connecting after the roadway portion. For example, at a turn, the measured travel path 160 of the vehicle 110 traveling straight may be (statistically) different from the measured travel path 160 of the vehicle 110 turning. This is exemplarily illustrated in FIG. 4a for a turning situation 400. In particular, FIG. 4a shows a (core) roadway portion 410 followed by a first (peripheral) roadway portion 410 (in straight driving) and followed by a second (peripheral) roadway portion 410 (when turning right).

[0113] In particular, it can be seen from Fig. 4a that the measured travel path 160 differs in the (core) roadway portion 410 before a turn, depending on whether the vehicle 110 drives straight ahead or turns following the (core) roadway portion 410. Thus, the measured travel path 160 in the (core) roadway portion 410 may depend on in which of one or more subsequent (surrounding) roadway portions 410 the respectively measured travel path 160 extends.

[0114] In a corresponding manner, the travel path 160 measured for a (core) roadway portion 410 may depend on one or more preceding (periphery) roadway portions 410 through which the respective measured travel path 160 extends.

[0115] The central unit 120 can be configured to determine, for one (core) roadway portion 410, a number of different reference driving paths 210 for different preceding and / or following (surrounding) roadway portions 410. In particular, for a given core roadway portion 410, all possible partial sequences of directly successive roadway portions 410 (e.g., based on data of a (SD) map) can be determined, which partial sequences have a predetermined number N of surrounding roadway portions 410 in front of the given core roadway portion 410 and / or a predetermined number M of surrounding roadway portions 410 behind the given core roadway portion 410. N and / or M may be 1 or more or 2 or more, respectively. Then, for each possible partial sequence of roadway portions 410 (each including a given core roadway portion 410), a reference driving path 210 in the given core roadway portion 410 can be determined. For a given core roadway portion 410, if there are different possible partial sequences of roadway portions 410, it is possible to identify a number Q of different reference driving paths 210 (e.g., Q is greater than 1 or greater than 2).

[0116] For this purpose, the central unit 120 can be configured to assign each of the available measured driving paths 160 for a given core roadway portion 410 to one of the Q sequences of roadway portions 410. Thus, in the example illustrated in FIG. 4a, the driving paths 160 measured in a straight direction are assigned to a first sequence of roadway portions 410, and the driving paths 160 measured in a right-turn direction are assigned to a second sequence of roadway portions 410.

[0117] Thus, for each sequence of roadway sections 410, a respective subset of the measured travel path 160 can be determined. Then, based on each subset of the measured travel path 160, a respective reference travel path 210 can be determined (by the method 300 described herein). Then, Q reference travel paths 210 can be provided as map data 125 for a given roadway section 410. Then, during the operation of the (automated) vehicle 110, a series of roadway sections 410 of the travel route of the vehicle 110 can be determined. Then, for the passage of a given core roadway section 410, a reference travel path 210 that matches this can be selected (and used for the passage of the core roadway section 410 to determine the target trajectory of the vehicle 110).

[0118] For the connection of the reference driving path 210 of the consecutive roadway portions 410, it may be advantageous to provide an interval between the end point of the reference driving path 210 of the roadway portion 410 and the end or end point 412 of the roadway portion 410 (as shown exemplarily in FIG. 4b). In a corresponding embodiment, it is also possible to provide an interval 420 between the start or start point 411 of the roadway portion 410 and the start or start point of the reference driving path 210 of the roadway portion 410. The interval 420 may correspond, for example, to half the interval between two consecutive planes 271. The start point of the reference driving path 210 may correspond to the first support point plane 271 and / or the end point of the reference driving path 210 may correspond to the support point 200 of the last support point plane 271.

[0119] As shown in Fig. 4b, the last support point 200 of the reference driving path 210 of the roadway section 410 has a certain distance 420 from the end 412 of the roadway section 410. Also, the first support point 200 of the reference driving path 210 of the following roadway section 410 has a certain distance 420 from the beginning 411 of the following roadway section 410. Then, in order to identify the connected reference driving paths 210 of the consecutive roadway sections 410, it is possible to fit a (possibly linear) connecting segment 421 between these two support points 200 of the reference driving path 210 of the roadway section 410.

[0120] As shown in Fig. 4a, at a branch in the road network, the natural driving paths 160 may differ, for example, between a straight-moving vehicular driver and a turning vehicular driver in the same driving lane. By the measures described herein, the determination of an incorrect reference driving path 210 can be reliably avoided by averaging different groups of the natural driving paths 160. Also, discontinuities between the reference driving paths 210 of successive roadway portions 410 can be avoided. In particular, the above-mentioned measures can provide a continuous transition between the reference driving paths 210 of successive roadway portions 410.

[0121] As described herein, the method 300 can be applied separately for each roadway portion 410 and the order of passing through the roadway portions 410, rather than for each roadway portion 410. In this case, a core roadway portion 410 can be considered (for which the measured travel path 160 is evaluated). Also, the order of the roadway portions 410 to be passed around the core roadway portion 410 can be considered. Here, the perimeter can be defined by the maximum total length of the roadway portions 410 joined together (in front of and / or behind the core roadway portion 410, respectively). The maximum total length can be, for example, 100 to 200 meters. The roadway portions 410 can have a length of 20 to 50 meters. Thus, it is possible to consider an order or sequence of roadway portions 410, each having four or more roadway portions 410.

[0122] To allow for a continuous transition between the travel paths 210 of the adjacent road sections 410, the planes 271 perpendicular to the road course 270 of the road section 410 may not be evaluated along the entire length of the road section 410, but may maintain a certain distance 420 at the edges 411, 412, for example half the distance between the two planes 271, respectively. In this way, when synthesizing the reference travel paths 210 of the individual road sections 410 into a consistent map of the road network, it is possible to fit continuous transitions 421 between the reference travel paths 210 of the individual road sections 410.

[0123] The fitting of the transition portion 421 can be performed such that the reference driving path 210 of the road portion 410 is extended in the driving direction by the first position point of the respective subsequent reference driving path 210 while recognizing the identified reference driving path 210 of the surrounding (especially adjacent) road portion 410. Thus, each reference driving path 210 can be extended individually and the process of fitting the transition portion 421 can be parallelized.

[0124] In this case, the set of consecutive reference driving paths 210 can be identified based on the geospatial distance metric (position and / or orientation) of each start point and end point of the reference driving paths 210. Alternatively or additionally, the Jaccard distance can be used as a criterion for the allocation of the reference driving paths 210 in the consecutive roadway portions 410. In this case, the Jaccard distance can be identified based on the measured trajectory 160 that identifies the potentially consecutive reference driving paths 210.

[0125] 5a shows a flowchart of a (possibly computer-implemented) method 500 for determining a reference driving path 210 for a core roadway portion 410 (i.e., for a roadway portion 410 referred to as the core roadway portion for purposes of a univocal description of the method). The method 500 includes determining 501 a plurality of measured driving paths 160 of one or more vehicles 110 for a corresponding plurality of traversals of the core roadway portion 410.

[0126] The method 500 further includes assigning 502 each of the measured driving paths 160 to a respective one of the sets of different sub-sequences to identify, for each sub-sequence in the set of sub-sequences, a respective one subset of the measured driving paths 160. In this case, each of the different sub-sequences may have a core roadway portion 410 and at least one surrounding roadway portion 410 located immediately before and / or after the core roadway portion 410 in the driving direction. Thus, each of the sub-sequences may have one or more roadway portions 410.

[0127] Additionally, the method 500 includes, for each sub-sequence in the set of sub-sequences, determining 503 a reference traveled path 210 for the core roadway portion 410 based on a respective subset of the measured traveled path 160. For this purpose, for example, the method 300 can be used.

[0128] 5b shows a flowchart of a (possibly computer-implemented) method 510 for at least partially automated longitudinal and / or lateral steering of a vehicle 110 along a travel route through a road network. The method 510 can be performed by a control unit 111 of the vehicle 110.

[0129] The method 510 includes determining 511 a set of different reference driving paths 210 for a corresponding set of different partial sequences for a previous passage of a core roadway portion 410 of the road network in the driving route, in particular based on map data of a digital map (e.g. an HD map) of the road network. Additionally, the method 510 includes identifying 512 a partial sequence of the set of different partial sequences based on the driving route. In particular, it is possible to determine a partial sequence that corresponds to the driving route, i.e. that extends along the driving route.

[0130] Further, the method 510 includes determining 513 a target trajectory for the vehicle 110 for at least partially automated longitudinal and / or lateral driving when passing through the core roadway portion 410 based on the reference driving path 210 for the identified partial sequence.

[0131] The determination of the reference driving paths 210 for a series of roadway sections 410 with a relatively large curvature (e.g. due to a turning situation) may lead to significant discontinuities at the transitions between the reference driving paths for the roadway sections 410 that directly follow each other. This is exemplarily illustrated in FIG. 6. In particular, FIG. 6 shows the support points 200 (each shown as a white circle) of the reference driving path 210 for a first roadway section 410 and the support points 200 (each shown as a hatched circle) of the reference driving path 210 for the subsequent second roadway section 410. In this case, the course 270 of the second roadway section 410 is essentially perpendicular to the course 270 of the first roadway section 410. This results in the support point plane 271 for the second roadway section 410 being arranged essentially perpendicular to the support point plane 271 for the first roadway section 410.

[0132] 6, a relatively large difference between the orientation of the support point plane 271 and the course of the measured travel path 160 may lead to an overlap and / or discontinuity between one or more support points 200 at the end 412 of the first roadway section 410 and one or more support points 200 at the beginning 411 of the subsequent second roadway section 410. This may again lead to a discontinuity at the transition between the nominal travel paths 210 of both roadway sections 410.

[0133] As exemplarily illustrated in FIG. 7 , the central unit 120 can be configured to determine a reference course 700 of the roadway portion 410 based on a measured travel path 160 of the roadway portion 410. For example, the measured travel path 160 (possibly without one or more outliers) can be determined to determine the reference course 700. The reference course 700 can then be used instead of the map course 270 of the roadway portion 410 to determine a series of support point planes 200 for the roadway portion 410. In this case, each of the support point planes 200 can be positioned perpendicular to the reference course 700.

[0134] 3 and 2a-2d can be used correspondingly to determine the support points 200 for the reference driving path 210. Thus, the accuracy of the determined reference driving path 210 can be improved, especially with respect to the transition to the reference driving path 210 for the subsequent roadway portion 410.

[0135] As can be seen from Fig. 7, the beginning 411 (i.e. the start point or the position of the start point) of the map course 270 of the roadway portion 410 for which the travel path 210 is to be determined may be different from the beginning 701 (i.e. the start point or the position of the start point) of the corresponding reference course 700. Alternatively or additionally, the end 412 (i.e. the end point or the position of the end point) of the map course 270 of the roadway portion 410 may be different from the end 702 (i.e. the end point or the position of the end point) of the corresponding reference course 700. The central unit 120 may be configured to determine the beginning 701 and / or the end 702 of the reference course 700 so as to reduce, in particular minimize, a predetermined distance criterion (spacing criterion) or clearance. The distance criterion may depend on the distance of each point 701, 702 of the reference course 700 from the corresponding point 411, 412 of the map course 270. Alternatively, or in addition, the distance criterion may depend on the deviation of the orientation 720 of the start 701 or end 702 of the reference course 700 from the orientation 720 of the start 411 or end 412 of the map course 270. Thus, the start 701 and / or end 702 of the reference course 700 can be specified such that its distance to the start 411 and / or end 412 of the map course 270 is as small as possible and / or such that the orientation 720 of the start 701 and / or end 702 of the reference course 700 is as close as possible to the orientation 720 of the start 411 and / or end 412 of the map course 270.

[0136] As described herein, the driving path 210 for each road portion 410 can be efficiently determined depending on the sequence of traversed road portions 410, i.e., depending on the turning direction.

[0137] When generating a driving path 210 from the fleet data 115 of multiple vehicles 110, the road geometry contained in the SD map, i.e. the map course 270, can be used to construct cross sections 271 perpendicular to the road course 270 in which piercing points 272 of the measured trajectory 160 of multiple passes of each road portion 410 are identified. To identify a representation of one or more reference driving paths 210, the piercing points 272 are grouped (into one or more support points 200) for each plane 271, and one or more support points 200 are connected (by path segments 211) between the planes 271.

[0138] Limitations in the depiction of road geometry and / or curvature progression in SD maps may lead to an inability to accurately learn turning paths at intersections because the constructed plane 271 only partially intersects the measured trajectory 160 of the passage and / or the ordering of the constructed plane 271 along the road geometry 270 results in intersections 272 with the measured trajectory 160 whose progression coordinates along the measured trajectory 160 are not monotonically increasing or decreasing.

[0139] This specification describes a method for determining road course geometry based on fleet data, which can distinguish between driving and turning directions at intersections and therefore can depict the actual curvature of the roadway 150 much better than SD maps.

[0140] The starting point is a collection of measured trajectories 160 of passes over the same (core) road portion 410. The collection of measured trajectories 160 may then refer to trajectories that all follow the same road portion 410 order around the core road portion 410 and / or all have the same turning direction (if intersections are located around the core road portion 410).

[0141] In a first step, the measured trajectories 160 of the passages can be adjusted so that all the trajectories 160 cover the same part 410 of the road network. For this purpose, each part 161 of the measured trajectories 160 can be assigned a road part 410 of the SD map according to a predefined criterion. In this process, called map matching, the local proximity, orientation or assignment of surrounding points can be used as the assignment criterion. The course coordinates are determined along the measured trajectory 160 for the last point before the reference point in the road network of the SD map and for the first (first) point after the reference point, and then, based on the course coordinates of the measured trajectory 160 and based on the course coordinates in the SD road network, the adjustment of the measured trajectory 160 can be performed so that the course coordinates of the measured trajectory 160 with which the measured trajectory 160 should be crossed are determined. Thus, based on the measurement data 115 of the vehicle 110, it is possible to determine the measured trajectory 160 of the passage of a given road part 410.

[0142] Advantageously, the adjustment is performed not only on the core road section 410 itself, but also on an extended range of a certain length (e.g. 15 meters) along a set of surrounding SD road sections, thus improving the robustness of the method.

[0143] For the measured tracks 160 to be adjusted for the (core) road section 410, a common relative path length can be defined, i.e. a path length for the entire length of the respective adjusted track 160 (for example 0%, 10%, 20%...100% of the respective length). At the corresponding points of the relative path lengths, a position can be determined for each track 160 and in particular can be interpolated. The thus determined positions of the same relative path lengths of all tracks 160 can then be summed up, on the basis of which a representative position for these relative path lengths can be determined. In this case, it is possible to use an average value formation for each coordinate. In this case, outliers are advantageously eliminated (for example elimination of 10% of the maximum and 10% of the minimum of all values ​​before the average value formation).

[0144] The sequence of positions thus identified can be interpreted as a polyline or preliminary step for the improved road path geometry. In this specification, the sequence is also referred to as the reference path 700 of the core road portion 410. Even if the trajectory 160 is adjusted to an extended range compared to the core road portion 410, it is possible to identify the start 701 and end 702 of the road path geometry assigned to the core road portion 410. This can be brought about, for example, by orthogonal projection of the start 411 and end 412 of the core road portion 410 onto the newly identified road path geometry 700.

[0145] Particularly during turning maneuvers (e.g. so-called U-turns), the determined transition point 702 to the following road section 410 may depend relatively strongly on the geometric position of the determined road course geometry 700. The end 702 of the road section 410 may then differ relatively greatly from the start 701 of the following section 410, resulting in a jump in the road course. To prevent this, it is possible to select points along the road course geometry 700 to be determined as start or end points 701, 702 which not only have the smallest possible distance to the corresponding points 411, 412 in the SD map, but also whose orientation 720 corresponds as well as possible to the corresponding points 411, 412. If the geometry in the SD map is depicted by a polyline, the orientation 720 of the points 411, 412 in the SD map can be determined as the average orientation of both adjacent line segments.

[0146] 8a shows a flowchart of a (possibly computer-implemented) method 800 for determining a reference driving path 210 for a roadway portion 410. The method 800 includes determining 801 a plurality of measured driving paths 160 of one or more vehicles 110 for a corresponding plurality of traverses of the roadway portion 410. The method 800 further includes determining 802 a reference course 700 for the roadway portion 410 based on the plurality of measured driving paths 160. The reference course 700 can then be used in the method 300 as a course 270, 700 to determine at least one reference driving path 210 for the roadway portion 410.

[0147] Thus, in particular, the method 800 may include locating 803 a series of support planes 271 along the reference path 700 and, for each of the plurality of measured travel paths 160, identifying 804 a series of intersections 272 of each measured travel path 160 with the corresponding series of support planes 271. Additionally, the method 800 may include identifying 805 at least one roadway portion 410 for the roadway portion 410 based on the plurality of identified series of intersections 272 for the corresponding plurality of measured travel paths 160.

[0148] 8b shows a flowchart of a (possibly computer-implemented) method 810 for identifying a reference trajectory 700 of a core roadway portion 410 (i.e., a roadway portion referred to as a core roadway portion). The method 810 includes identifying 811 a plurality of measured travel paths 160 of one or more vehicles 110 for a corresponding plurality of traversals of the core roadway portion 410.

[0149] The method 810 further includes assigning 812 a first subset of the measured driving paths 160 of the plurality of measured driving paths 160 to a first subset of the set of distinct subsets, where each distinct subset may include a core roadway portion 410 and at least one surrounding roadway portion 410 located immediately before and / or immediately after the core roadway portion 410 in the driving direction.

[0150] The method 810 further includes identifying 813 a first reference path 710 of the core roadway portion 410 for the first sub-sequence based on (only) a first subset of the measured traveled path 160 for the first sub-sequence. Then, as herein, the first reference path 700 can be identified based on (only) a first subset of the measured traveled path 160.

[0151] Thus, it is already possible to take into account from which surrounding roadway portion 700 the vehicle 100 is traveling into the core roadway portion 700 and / or into which surrounding roadway portion 700 the vehicle is traveling out of the core roadway portion 700 when determining the reference path 700 for the core roadway portion 700. It is thus possible to determine a reference path 700 that depends on the direction of travel. Thus, the quality of the determined reference travel path 210 can be further improved.

[0152] In particular, different reference driving paths 210 for different partial sequences can be identified using the specific reference progression 700 for each partial sequence, respectively. Therefore, the quality of the identified reference driving paths 210 can be significantly improved.

[0153] The reference trajectory 700 of the core roadway portion 410 for a given sub-sequence can be provided as map data for a digital map, and the reference trajectory 700 can be displayed (e.g., as an augmented reality display) on a head-up display of the vehicle 110, for example, as the core roadway portion 410 is traversed.

[0154] The measures described herein allow efficient and accurate identification of a reference driving path 210 for a roadway portion 410 of a road network, which can be provided as map data 125 for an HD map. The identified reference driving path 210 can be used by the at least partially autonomously operating vehicle 110 to identify a target trajectory for the vehicle 110. Thus, the quality of the autonomously operating vehicle 110 can be improved.

[0155] It should be noted that the invention is not limited to the embodiments shown, and in particular that the specification and drawings are merely illustrative of the principles of the proposed methods, devices and systems.

Claims

1. An apparatus (120) for identifying a reference course (700) of a core roadway section (410), the apparatus (120) comprising: - identifying a plurality of measured travel paths (160) of one or more vehicles (110) for a corresponding plurality of traverses of said core roadway portion (410); - assigning a first subset of the measured driving paths (160) of the plurality of measured driving paths (160) to a first sub-sequence of a set of different sub-sequences, each of the different sub-sequences having the core roadway portion (410) and at least one surrounding roadway portion (410) located immediately before and / or immediately after the core roadway portion (410) in the direction of travel; and - determining a first reference course (700) of the core roadway portion (410) for the first sub-sequence based on the first subset of measured traveled paths (160) for the first sub-sequence; 1. An apparatus (120) comprising:

2. The device (120) - identifying one or more measured traveled paths (160) of a first subset of measured traveled paths (160) as outliers; and - not to take into account one or more of the identified and measured travel paths when determining the first reference course (700) of the core roadway portion (410); 2. The apparatus (120) of claim 1, wherein the apparatus (120) is configured to:

3. The device (120) - determining a total length for each of one or more measured travel paths (160) of said first subset of measured travel paths (160), - for one or more measured travel paths (160) of the first subset of measured travel paths (160), respectively, identifying a series of path points at a corresponding series of successive extended positions, each successive extended position being relative to the total length of the measured travel path (160); and - for each of said extended positions, determining a corresponding path point of the first reference course (700) based on each of said path points of the measured travel path (160), in particular based on an average value, for example a trimmed average value, of said path points of the measured travel path (160), in order to determine a series of path points of the first reference course (700) at a corresponding series of consecutive said extended positions of the first reference course (700); 3. The device (120) of claim 1 or 2, characterized in that it is configured

4. The device (120) for one extension position in a series of successive extension positions: - identifying one or more path points of said measured traveled path (160) as outliers; and - one or more of the identified path points of the measured travel path (160) are not taken into account when determining the corresponding path points of the first reference course (700), 4. The apparatus (120) of claim 3, wherein the apparatus is configured to:

5. The device (120) calculates, for a segment of the first reference path (700) between a first path point and a second path point of the first reference path (700), in particular: - a first speed value is determined for one or more corresponding first path points of a first subset of the measured traveled path (160) that identified the first path points of the first reference path (700), - determining second speed values ​​for one or more corresponding second path points of a first subset of the measured traveled path (160) that identified second path points of the first reference path (700); and - based on said first and second velocity values, in particular based on an average value of said first and / or second velocity values, an empirical velocity is determined for said segment of said first reference path (700), The apparatus (120) of claim 3, configured to identify an empirical velocity.

6. The device (120) - based on a digital map, for said core roadway portion (410), identifying a map course (270) of said core roadway portion (410) shown on said digital map; and - identifying said first reference course (700) of said core roadway portion (410) also based on said map course (270); 3. The device (120) of claim 1 or 2, characterized in that it is configured

7. The device (120) - to identify the beginning (411) and / or the end (412) of said map course (270), and - to identify the corresponding beginning (701) and / or the corresponding end (702) of said first reference course (700) based on the clearance; It is composed of the clearance for identifying the beginning (701) of the first reference course (700) depends on the distance between the beginning (411) of the map course (270) and the beginning (701) of the reference course (700) to be identified; and / or the clearance for identifying the end (702) of the first reference course (700) depends on the distance between the end (412) of the map course (270) and the end (702) of the reference course (700) to be identified; 7. The apparatus (120) of claim 6.

8. the clearance for identifying the beginning (701) of the first reference course (700) depends on the deviation between the orientation (720) of the map course (270) at the beginning (411) of the map course (270) and the orientation (720) of the first reference course (700) at the beginning (701) of the first reference course (700) to be identified; and / or the clearance for identifying the end (702) of the first reference course (700) depends on the deviation between the orientation (720) of the map course (270) at the end (412) of the map course (270) and the orientation (720) of the first reference course (700) at the end (702) of the first reference course (700) to be identified; 8. The device (120) of claim 7.

9. The device (120) - to arrange a series of support point planes along said first reference path (700), - for each measured travel path (160) of the first subset of measured travel paths (160), determining a respective series of intersections (272) of said measured travel path (160) with a corresponding series of said support point planes (271); and - determining, based on the determined sequence of intersections (272) for the first subset of the measured traveled paths (160), at least one reference traveled path (210) for the core roadway portion (410) for the first subset.

3. The device (120) of claim 1 or 2, characterized in that it is configured

10. - the first partial sequence comprises the core roadway portion and a predetermined peripheral roadway portion (410) located immediately after or immediately before the core roadway portion (410); and said device (120) - determining a reference driving path (210) for the core roadway portion (410) based on the first subset of measured driving paths (160) and using the first reference path (700); - to identify a preceding or succeeding reference driving path (210) for a given said surrounding roadway portion (410); and - combining the reference driving path (210) for the core roadway portion (410) with the subsequent or preceding reference driving path (210) for a given surrounding roadway portion (410) to identify a sequence driving path for the first partial sequence; 3. The device (120) of claim 1 or 2, characterized in that it is configured

11. The device (120) of claim 1 or 2, characterized in that the device (120) is configured to operate the vehicle (110) in dependence on the identified first reference path (700) of the core roadway portion (410) when traveling through the core roadway portion (410).

12. The device (120) according to claim 1 or 2, characterized in that the device (120) provides the first reference course (700) for the digital map as map data (125) relating to the core roadway portion (410), in particular together with an empirical speed progression along the first reference course (700).

13. A device (120) for identifying a reference driving path (210) for a roadway portion (410), the device (120) comprising: - identifying a plurality of measured travel paths (160) of one or more vehicles (110) for a corresponding plurality of traverses of said roadway portion (410); - determining a reference course (700) of said roadway section (410) based on a plurality of said measured travel paths (160), - to place a series of support point planes (271) along said reference path (700), - for each of said plurality of measured travel paths (160), determining a respective series of intersections (272) of said respective measured travel paths (160) with a corresponding series of said support point planes (271); and - determining, based on a plurality of measured series of said intersections (272), for a corresponding plurality of measured travel paths (160), at least one reference travel path (210) for said roadway portion (410); 1. An apparatus (120) comprising:

14. A method (810) for identifying a reference course (700) of a core roadway section (410), the method (810) comprising: - identifying (811) a plurality of measured travel paths (160) of one or more vehicles (110) for a corresponding plurality of traverses of said core roadway portion (410); - assigning (812) a first subset of the measured driving paths (160) from the plurality of measured driving paths (160) to a first sub-sequence from a set of different sub-sequences, each of the different sub-sequences comprising the core roadway portion (410) and at least one surrounding roadway portion (410) located immediately before and / or immediately after the core roadway portion (410) in the direction of travel; and - determining (813) a first reference course (700) of the core roadway portion (410) for the first sub-sequence based on the first subset of measured traveled paths (160) for the first sub-sequence; 8. A method (810) comprising:

15. A method (800) for identifying a reference driving path (210) for a roadway portion (410), the method (800) comprising: - determining (801) a plurality of measured travel paths (160) of one or more vehicles (110) for a corresponding plurality of traverses of said roadway portion (410); - determining (802) a reference course (700) of said roadway section (410) based on a plurality of said measured travel paths (160); - placing (803) a series of support point planes (271) along said reference path (700); - for each of the plurality of measured travel paths (160), identifying (804) a respective series of intersections (272) of each measured travel path (160) with a corresponding series of support point planes (271); and - determining (805) at least one reference travel path (210) for the roadway portion (410) for a corresponding plurality of measured travel paths (160) based on a plurality of measured series of the intersection points (272); 8. A method (800) comprising:

Images