METHOD FOR ASCERTAINING AN ATTRIBUTE ON A DIGITAL MAP FOR A VEHICLE

Abstract

A method for assigning a map attribute in an digital map for a vehicle. The method includes: ascertaining a validity range of the map attribute stored in absolute coordinates in relation to a reference line of the digital map; at least one base point of a defined geometry being ascertained originating from the map attribute using a defined segment of the reference line of the vehicle; and the map attribute being assigned to the defined segment intersected at the base point.

Claims

1. A method for assigning a map attribute in a digital map for a vehicle, comprising the following steps: ascertaining a validity range of the map attribute stored in absolute coordinates in relation to a reference line of the digital map; ascertaining at least one base point of a defined geometry originating from the map attribute using a defined segment of the reference line; and assigning the map attribute to the defined segment intersected at the base point.

2. The method as recited in claim 1, wherein a position of the map attribute is ascertained on a segment of the reference line.

3. The method as recited in claim 2, wherein an offset of the map attribute on the reference line is ascertained from the ascertained position of the map attribute.

4. The method as recited in claim 1, wherein the intersection point is ascertained using concentric circles which become larger in a defined manner, or using defined rays originating from the map attribute.

5. The method as recited in claim 1, wherein a lane center of a roadway is used as the reference line.

6. The method as recited in claim 1, wherein for the case in which two segments are intersected, lengths of the intersected segments are ascertained, and the map attribute is assigned to the segment having the greatest distance.

7. The method as recited in claim 1, wherein in the case in which two segments of the reference line have the same distance to the map attribute, the map attribute is assigned to that segment of the two segments in a travel direction of the vehicle.

8. The method as recited in claim 1, wherein the method is carried out during creation of the digital map, on a control unit or in the cloud.

9. A device configured to assign a map attribute in a digital map for a vehicle, the device configured to: ascertain a validity range of the map attribute stored in absolute coordinates in relation to a reference line of the digital map; ascertain at least one base point of a defined geometry originating from the map attribute using a defined segment of the reference line; and assign the map attribute to the defined segment intersected at the base point.

10. A non-transitory computer-readable data carrier on which is stored a computer program for assigning a map attribute in a digital map for a vehicle, the computer program, when executed by a computer, causing the computer to perform the following steps: ascertaining a validity range of the map attribute stored in absolute coordinates in relation to a reference line of the digital map; ascertaining at least one base point of a defined geometry originating from the map attribute using a defined segment of the reference line; and assigning the map attribute to the defined segment intersected at the base point.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0026] FIG. 1 shows a representation of fundamental features of the provided method of an example embodiment of the present invention.

[0027] FIGS. 2A-2E show an explanation of the conventional nearest neighbor method.

[0028] FIG. 3A-3D show a schematic representation of a mode of operation of the provided method in accordance with an example embodiment of the present invention.

[0029] FIGS. 4A-4D show a schematic representation of another specific example embodiment of the provided method in accordance with the present invention.

[0030] FIGS. 5A and 5B show a schematic representation of another specific example embodiment of the provided method, in accordance with the present invention.

[0031] FIGS. 6A and 6B show a schematic representation of another specific example embodiment of the provided method in accordance with the present invention.

[0032] FIG. 7 shows a schematic sequence of a specific example embodiment of the provided method in accordance with the present invention

DETAILED DESCRIPTION OF EXAMPLE EMBODIMENTS

[0033] The term “automated vehicle” is used synonymously hereinafter with the terms “fully automated vehicle,” “autonomous vehicle,” “semiautonomous vehicle,” “E-vehicle,” and other driver assistance functions.

[0034] FIG. 1 explains fundamental features of the provided method.

[0035] An offset O (for example, a distance from a defined position of the vehicle on a reference line including multiple strung together segments S.sub.1 . . . S.sub.n) of a map attribute A may not be read out directly from a high-resolution digital map, but rather has to be ascertained by computer based on a reference line representing the road geometry (for example, in the form of a lane reference line) and the position of map attribute A stored in absolute coordinates in the map. Offset O advantageously does not necessarily have to be ascertained beginning from the vehicle position, but rather may be ascertained beginning from an arbitrarily defined point. A geometry of the reference line is recognizable including shape points P.sub.1 . . . P.sub.n, whose connections are referred to as segments S.sub.1 . . . S.sub.n, and a map attribute A, which is mapped orthogonally on the reference line, for example, to create an electronic horizon for the vehicle (not shown), in order to thus ascertain a base point f.sub.p and therefrom offset O (distance of shape point P.sub.1 to base point f.sub.p).

[0036] A conventional algorithm for determining offset O may be expressed as follows in pseudocode: [0037] 1. Iterate over all segments S.sub.i [0038] 2. Calculate base point f.sub.p between map attribute A and segment S.sub.i and calculate distance d between the two points [0039] 3. If distance d is less than distance d ascertained in the prior iteration, overwrite distance d and also store base point f.sub.p [0040] 4. After the end of the loop over all segments S.sub.i, nearest base point f.sub.p is known [0041] 5. Calculate offset O as distance d between shape point P.sub.1 and base point f.sub.p along the segments

[0042] This algorithm, which is known per se, may be inefficient, however, since it is necessary to iterate over all segments S.sub.1 . . . S.sub.n of the geometry. A so-called “nearest neighbor method,” which is known per se, is therefore often used, which may significantly reduce the segments coming into consideration, using which a base point f.sub.p has to be calculated.

[0043] FIGS. 2A-2E outline, in five images A through E, a mapping according to the conventional nearest neighbor approach, which may be expressed as follows in pseudocode: [0044] 1. Calculate distances a.sub.i between map attribute A and shape points P.sub.i (FIGS. 2A, 2B) [0045] 2. Select from these distances a.sub.i shape point P.sub.i of the reference line having the least distance and use exclusively segments S.sub.i-1 and S.sub.i associated with this shape point to calculate base point f.sub.p (FIG. 2C) [0046] 3. Calculate base point f.sub.p in which map attribute A is orthogonally mapped on the segment (FIG. 2D) [0047] 4. Calculate offset O (FIG. 2E)

[0048] The explained conventional nearest neighbor method is not capable, however, in certain situations, for example, on very curvy roads, as occur in urban areas, or on serpentines in the mountains, under certain circumstances of calculating the correct mapping of map attribute A, because mapping is carried out on incorrect segments.

[0049] A method is therefore provided for improved calculation of offsets O in relation to map attributes A in the supply of an electronic horizon for control units (for example, in autonomous or semi-autonomous motor vehicles, E-vehicles, etc.).

[0050] The background of the provided method is that offset O may not be read out directly from the digital map, but rather has to be ascertained by computer based on a reference line representing the road geometry and the position of map attribute A (for example, tunnel, speed limits, etc.). The geometry of the reference line is made up of shape points P.sub.1 . . . P.sub.n, whose connections are called segments S.sub.0 . . . S.sub.n.

[0051] Map attribute A is mapped orthogonally on this geometry to create the electronic horizon. Various procedures or algorithms are possible for calculating offset O, the above-explained conventional methods being inefficient or possibly supplying incorrect results in certain situations.

[0052] In particular, the conventional nearest neighbor method may calculate incorrect offsets O on sharply curved roads or serpentines. In the nearest neighbor method, firstly the distances between map attribute A and the shape points are calculated, only the shape point having the least distance being used for calculating base point f.sub.p orthogonally on the corresponding segment. It is apparent in FIG. 3A that distance a1 is less than distances a2 and a3, in this case map attribute A would be mapped on an “incorrect” segment of the reference line.

[0053] The provided method avoids the disadvantages of the nearest neighbor method in that, for example, a concentric circle around map attribute A is enlarged until it intersects a segment as indicated in principle in one specific embodiment of the provided method in FIGS. 3A-3D using four images A through D.

[0054] The segment intersected by the concentric circle is subsequently used as the basis for the calculation of base point f.sub.p and offset O. If the concentric circle should intersect multiple segments S.sub.1 . . . S.sub.n, nonetheless the nearest segment may be determined. For this purpose, for each of the intersected segments, the distance of the two intersection points with the circle has to be determined. That segment in which the distance is greatest is nearest to map attribute A and is selected. It is indicated in FIG. 5A that segment S3 is the one at which distance a1 is greater than distance a2. Therefore, segment S3, which is intersected by the circle, is selected and assigned to map attribute A at base point f.sub.p, as indicated in FIG. 5B.

[0055] Alternatively to the use of concentric circles, the use of two-dimensional ray casting is also possible. In this ray casting approach, rays are drawn originating from map attribute A and intersected with segments, nearest segment including base point f.sub.p and offset O being able to be determined from their distance to the segments.

[0056] Several map attributes A are listed solely as examples hereinafter, which may be mapped on the reference line: [0057] Fixed map attributes, for example, tunnels on roads [0058] Flexible map attributes, for example, speed limits on roads [0059] Lane groups including their lanes and their lane markings [0060] Landmarks or locating objects [0061] Referencing objects for updating attributes [0062] Charging stations (for E-vehicles) [0063] Devices for calculating the range using present battery charge on the basis of the road condition (for E-vehicles)

[0064] Map attributes A which are mentioned and others which are not mentioned occur in large numbers in the digital map, so that the mapping operation including the following improvement is often used. For a 1 km long horizon, for example, there may be approximately 1000 to 2000 map attributes for which the corresponding mapping calculations have to be carried out.

[0065] The procedure in the described method is indicated in FIGS. 3A-3D and 4A-4D and shows how it may solve the problem cases of the nearest neighbor method. In particular in the scenario of FIGS. 4A-4D, the conventional nearest neighbor method would probably supply incorrect results.

[0066] The algorithm operates in principle as follows: [0067] 1. Draw concentric circles around map attribute A which become larger until they intersect a segment of the reference line [0068] 2. Use the identified segment for the orthogonal mapping and calculate base point f.sub.p [0069] 3. Calculate offset O on the reference line

[0070] The last two steps do not differ from those of the known nearest neighbor method. The advantage over the nearest neighbor method results from the first step, since the testing by the intersection with the concentric circles relates directly to the segments and not initially only to the shape points.

[0071] FIGS. 5A and 5B explains a variant of the provided method:

[0072] If the concentric circle originating from map attribute A in step 1 should intersect multiple segments at the same time, the nearest segment may nonetheless be determined, as indicated in FIG. 5B. For this purpose, for each of the intersected segments, the distance of the two intersection points with the circle has to be determined. That segment in which the distance is greatest is nearest to map attribute A and is selected for base point f.sub.p. In the case of FIG. 5A, segment a1>a2, because of which map attribute A is mapped on segment S.sub.1 in FIG. 5B.

[0073] An alternative approach (not shown in the figures) to find the nearest segment in step 1 is the use of two-dimensional ray casting instead of the above-described concentric circles. In this ray casting approach (similarly to corresponding approaches in computer graphics), rays would originate starting from map attribute A, by whose distance to the segments the nearest segment of the reference line intersected by a ray may be determined.

[0074] In case of an equidistance of tangential segments to map attribute A, a travel direction R may advantageously be taken into consideration. In this case, that segment is selected on which the vehicle will soon be located or on which the vehicle will soon travel.

[0075] This scenario is indicated in FIG. 6A, where a travel direction R of the vehicle is indicated. It is apparent in FIG. 6B that due to the equidistance of map attribute A to intersected segments S1, S3, map attribute A is assigned due to travel direction R of the vehicle to segment S1, because the vehicle will travel earlier on segment S1 than segment S3. An uncertainty value ε is also taken into consideration as the equidistance, i.e., a=b+ε.

[0076] FIG. 7 very schematically shows a schematic sequence of a provided method.

[0077] In a step 100, a validity range of map attribute A stored in absolute coordinates is ascertained in relation to a reference line of the digital map.

[0078] In a step 110, at least one intersection point of a defined geometry originating from map attribute A with a defined segment S.sub.1 . . . S.sub.n of the reference line of the vehicle is ascertained.

[0079] In a step 120, map attribute A is assigned to intersected defined segment S.sub.1 . . . S.sub.n.

[0080] The provided method may advantageously be carried out locally in a control unit in the vehicle or centrally in the cloud. When the method is carried out, for example, a travel direction may also be taken into consideration, so that segments of the digital map including base point fp [sic; f.sub.p] already previously ascertained are transmitted to the vehicle.

[0081] The provided method may advantageously be implemented as software, which runs, for example, decentrally on a control unit in the vehicle or centrally in the cloud. Simple adaptability of the method is assisted in this way.

[0082] The provided method may advantageously be carried out in the cloud, the calculations in the cloud advantageously only being carried out a single time (for example, until the next map update). In the vehicle, the method is carried out during each trip, because in general highly updated, most recent map attributes are requested.

[0083] Those skilled in the art will modify the features of the present invention in a suitable way and/or combine them with one another, without departing from the core concept of the present invention.