Motion planning of an autonomous vehicle, motion planning system and vehicle with a motion planning system

1 . A method for motion planning of an autonomous vehicle, performed by a motion planning system, the method comprising:

determining a plurality of configuration samples through a configuration space using a sampling-based algorithm,

wherein the configuration space represents surroundings of the vehicle,

wherein the configuration samples are determined in predefined sampling steps,

wherein a sequence of the configuration samples defines a possible trajectory or path of the vehicle in the configuration space, and

wherein the configuration samples in each of the sampling steps are determined in parallel by using at least one graphics processing unit of the motion planning system;

calculating sequences of the configuration samples from the configuration samples by predefined motion criteria;

receiving a current vehicle position determined by a positioning sensor of the vehicle;

receiving route information from a navigation system of the vehicle based on the current vehicle position, wherein the route information is different from the current vehicle position;

including the route information in the configuration space or in a calculation by the predefined motion criteria,

wherein the predefined motion criteria include at least one of a distance to a target, a distance to an object, a velocity to reach the target, or a fastest way to a parking position;

outputting the sequences of the configuration samples as vehicle trajectories for moving the vehicle in the surroundings of the vehicle;

rasterizing the configuration space based on environmental information of the vehicle, wherein the configuration space is divided into a plurality of grid fields that form a rasterized configuration space, and wherein the environmental information of the vehicle includes information regarding a surface on a road or beside the road; and

after each of the sampling steps, determining a rating for each grid field of the plurality of grid fields of the rasterized configuration space according to environmental criteria,

wherein each grid field of the grid fields comprises two or more of the configuration samples, and

wherein each of the two or more of the configuration samples comprising each grid field of the grid fields receives the rating determined for the grid field according to the environmental criteria.

2 . The method according to claim 1 , further comprising:

after each of the sampling steps, clustering one or more of the sequences of the configuration samples with a same end state, wherein only one sequence of the one or more of the sequences of the configuration samples is processed further in the subsequent ones of the sampling steps, and wherein the one sequence of the one or more of the sequences of the configuration samples is selected by the predefined motion criteria.

3 . The method according to claim 1 , wherein the environmental criteria includes a classification of the surface and a Euclidian distance to a road limitation.

4 . The method according to claim 1 , further comprising:

after the vehicle trajectories are output, selecting a maneuver for the motion of the vehicle, wherein the maneuver comprises an overtaking, parking and/or breaking of the vehicle; and

controlling the vehicle to perform the maneuver.

5 . The method according to claim 1 , further comprising:

receiving static or dynamic object information from a vehicle sensor; and

including the static or dynamic object information in the configuration space or in a calculation by the predefined motion criteria.

6 . The method according to claim 1 , wherein the calculating the sequences of the configuration samples with the predefined motion criteria includes optimizing a cost function.

7 . A motion planning system comprising:

at least one processor; and

at least on memory storing instructions that, when executed by the at least one processor causes the motion planning system to:

determine a plurality of configuration samples through a configuration space using a sampling-based algorithm,

wherein the configuration space represents surroundings of a vehicle,

wherein the configuration samples are determined in predefined sampling steps,

wherein a sequence of the configuration samples defines a possible trajectory or path of the vehicle in the configuration space, and

wherein the configuration samples in each of the sampling steps are determined in parallel by using at least one graphics processing unit of the motion planning system;

calculate sequences of the configuration samples from the configuration samples by predefined motion criteria;

receive a current vehicle position determined by a positioning sensor of the vehicle;

receive route information from a navigation system of the vehicle based on the current vehicle position, wherein the route information is different from the current vehicle position;

include the route information in the configuration space or in a calculation by the predefined motion criteria,

wherein the predefined motion criteria include at least one of a distance to a target, a distance to an object, a velocity to reach the target, or a fastest way to a parking position;

output the sequences of the configuration samples as vehicle trajectories for moving the vehicle in the surroundings of the vehicle;

rasterize the configuration space based on environmental information of the vehicle, wherein the configuration space is divided into a plurality of grid fields that form a rasterized configuration space, and wherein the environmental information of the vehicle includes information regarding a surface on a road or beside the road; and

after each of the sampling steps, determine a rating for each grid field of the plurality of grid fields of the rasterized configuration space according to environmental criteria,

wherein each grid field of the grid fields comprises two or more of the configuration samples, and

wherein each of the two or more of the configuration samples comprising each grid field of the grid field receives the rating determined for the grid field according to the environmental criteria.

8 . A vehicle with the motion planning system according to claim 7 .

9 . The motion planning system according to claim 7 , wherein the environmental information of the vehicle includes information regarding a texture or a condition of the surface on the road or beside the road.

10 . The method according to claim 1 , wherein the environmental information of the vehicle includes information regarding a texture or a condition of the surface on the road or beside the road.