Method for the assessment of possible trajectories

1. A method for assessing possible trajectories of road users in a traffic environment comprising:

capturing the traffic environment with static and dynamic features;

identifying at least one first road user of the road users in the captured traffic environment;

determining at least one possible trajectory for the identified at least one first road user of the road users in the captured traffic environment; and

assessing the at least one determined possible trajectory for the identified at least one first road user of the road users based upon a behavior of a previously tracked at least one second road user of the road users in at least one previous traffic environment by using an adapted/trained recommendation service and the captured traffic environment.

2. The method according to claim 1 , wherein:

the adapted/trained recommendation service is configured to assess the at least one determined possible trajectory using a large number of corresponding combinations of assessed observed trajectories of the at least one second road user; and

the at least one previous traffic environment comprises different traffic environments.

3. The method according to claim 2 , wherein:

the adapted/trained recommendation service is a collaborative recommendation service and has a neural network, and

the neural network has a first autoencoder configured to capture the traffic environment and a second autoencoder configured to capture the at least one possible trajectory.

4. The method according to claim 2 , wherein:

the adapted/trained recommendation service is a collaborative recommendation service and is based on a K-nearest neighbor method, and

vectors of the K-nearest neighbor method are formed according to the large number of corresponding combinations of the assessed observed trajectories, and the different traffic environments.

5. The method according to claim 1 , wherein the adapted/trained recommendation service is a collaborative recommendation service and has a neural network with at least one convolution layer or a recursive neural network or is based on a K-nearest neighbor method.

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

determining the at least one possible trajectory using a geographical map of the traffic environment.

7. The method according to claim 1 , wherein capturing the traffic environment further comprises:

transforming spatial parts of the traffic environment into a two-dimensional reference system, corresponding to a plan view of the traffic environment.

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

determining the at least one possible trajectory according to an optimization of cost functions, a search-based method, and/or a machine-learning method.

9. A method for generating a recommendation service for assessing possible trajectories of a future road user in a future traffic environment based upon behavior of at least one tracked road user in at least one traffic environment, comprising:

determining a large number of corresponding combinations of captured traffic environments including the at least one traffic environment, observed trajectories, and possible trajectories, by repeatedly

capturing one of the traffic environments with at least one static feature and at least one dynamic feature,

identifying at least one road user of the at least one tracked road user in a traffic situation in the captured one of the traffic environments,

determining at least one possible trajectory for the identified at least one road user in the captured one of the captured traffic environments, and

capturing an observed trajectory associated with the determined at least one possible trajectory for the identified at least one road user; and

adapting the recommendation service based upon the large number of corresponding combinations of the captured traffic environments including the at least one traffic environment, the observed trajectories, and the possible trajectories, using a respective deviation of each of the captured observed trajectories from the associated determined at least one possible trajectory.

10. The method according to claim 9 , wherein at least one of the respective deviations is calculated by using a metric.

11. A method for planning a first trajectory to be driven by a first road user of a first traffic environment, comprising:

observing a plurality of second trajectories of a plurality of second road users in at least one second traffic environment including first static features;

training a recommendation service with the observed plurality of second trajectories;

capturing the first traffic environment, the first traffic environment including the first static features;

identifying the first road user in the captured first traffic environment;

identifying at least one third road user in the captured first traffic environment;

determining at least one possible third trajectory for the at least one third road user in the captured first traffic environment;

assessing the at least one determined possible third trajectory for the at least one third road user based upon the observed plurality of second trajectories by using the trained recommendation service and the captured first traffic environment; and

determining the planned first trajectory to be driven by the first road user in the first traffic environment based upon the assessment of the at least one possible third trajectory of the at least one third road user.

12. The method according to claim 11 , wherein an apparatus is set up to perform the method.

13. The method according to claim 11 , wherein a computer program, comprising commands which, when a computer executes the program, causes the computer to perform the method.

14. The method according to claim 13 , wherein the computer program is stored on a machine-readable storage medium.

15. The method according to claim 11 , further comprising:

adapting the trained recommendation service based upon an actual trajectory driven by the first user in the first environment after determining the first trajectory to be driven by the first road user in the first traffic environment.

16. The method according to claim 15 , further comprising:

calculating a difference between the actual trajectory and the determined first trajectory using a difference between an actual waypoint associated with the actual trajectory and a determined waypoint associated with the determined first trajectory.