Vehicle trajectory determination

• US 9,934,688 B2
• Filed: 07/31/2015
• Issued: 04/03/2018
• Est. Priority Date: 07/31/2015
• Status: Active Grant

• Current Assignee: Ford Global Technologies, LLC (Ford Motor Company), Board of Regents of the University of Michigan (University of Michigan), University of Michigan
• Original Assignee: Ford Global Technologies, LLC (Ford Motor Company), Board of Regents of the University of Michigan (University of Michigan)
• Inventors: Olson, Edwin, Galceran, Enric, Cunningham, Alexander G., Eustice, Ryan M., McBride, James Robert
• Primary Examiner(s): Mott, Genna M
• Application Number:
  US14/814,766
• Publication Number:
  US 20170031361A1
• Time in Patent Office: 977 Days
• Field of Search: None
• CPC Class Codes: B60W 2050/0029 Mathematical model of the driver
• CPC Class Codes: B60W 2050/0031 Mathematical model of the vehicle
• CPC Class Codes: B60W 2050/006 Interpolation; Extrapolation
• CPC Class Codes: B60W 2554/80 Spatial relation or speed relative to objects
• CPC Class Codes: B60W 2556/10 Historical data
• CPC Class Codes: B60W 2556/50 of positioning data, e.g. GPS [Global Positioning System] data
• CPC Class Codes: B60W 2556/65 Data transmitted between vehicles
• CPC Class Codes: B60W 30/08 Active safety systems predicting or avoiding probable or impending collision or attempting to minimise its consequences
• CPC Class Codes: B60W 30/0953 the prediction being responsive to vehicle dynamic parameters
• CPC Class Codes: B60W 30/0956 the prediction being responsive to traffic or environmental parameters
• CPC Class Codes: B60W 40/04 Traffic conditions
• CPC Class Codes: B60W 40/10 related to vehicle motion
• CPC Class Codes: B60W 50/0097 Predicting future conditions
• CPC Class Codes: G06V 20/56 exterior to a vehicle by using sensors mounted on the vehicle
• CPC Class Codes: G08G 1/163 involving continuous checking
• CPC Class Codes: G08G 1/166 for active traffic, e.g. moving vehicles, pedestrians, bikes
• CPC Class Codes: G08G 1/167 Driving aids for lane monitoring, lane changing, e.g. blind spot detection

Abstract

A system includes a computer programmed to identify, from a first vehicle, one or more second vehicles within a specified distance to the first vehicle. The computer is further programmed to receive data about operations of each of the second vehicles, including trajectory data. Based on the data, the computer is programmed to identify, for each of the second vehicles, a distribution of probabilities of each of a set of potential planned trajectories. The computer is further programmed to determine a planned trajectory for the first vehicle, based on the respective distributions of probabilities of each of the set of potential planned trajectories for each of the second vehicles. The computer is further programmed to provide an instruction to at least one controller associated with the first vehicle based on the determined planned trajectory.

Claims

1. A system comprising a computer programmed to:

receive data about a current and historical trajectory of each of the one or more second vehicles;

select from a set of three or more pre-defined policies, a set of potential policies for each second vehicle, each of the potential policies in each respective set of potential policies including a respective possible second vehicle trajectory and one or more rules for the respective second vehicle trajectory;

identify a set of change points for each second vehicle'"'"'s historical trajectory, wherein each change point specifies a location and a time along the respective second vehicle historical trajectory at which it is determined that the respective second vehicle changed from a first policy from the set of potential policies for the respective second vehicle to a second policy from the set of potential policies for the respective second vehicle;

determine, for each segment between a respective, adjacent pair of change points along the respective historical trajectory for each second vehicle, a respective policy from the set of potential policies for the respective second vehicle that fits to the historical trajectory of the respective second vehicle;

identify, based at least in part on the policy from the set of potential policies for each segment between the respective adjacent pair of change points that fits to the historical trajectory of each respective second vehicle, a distribution of probabilities for each second vehicle at a current time;

each probability of each respective distribution of probabilities associated with one of the potential policies from the set of potential policies for the respective second vehicle;

select a policy for the first vehicle from the set of three or more pre-defined policies, based at least in part on the respective distributions of probabilities at the current time for the respective one or more second vehicles; and

control the vehicle via an instruction provided by the computer to a controller based on the selected policy for the first vehicle.

2. The system of claim 1, wherein the selected policy for the first vehicle further indicates one or more alternative trajectories, each of the alternative trajectories associated with potential policies of the one or more second vehicles.

3. The system of claim 1, wherein the computer is further programmed to:

identify one or more candidate policies for the first vehicle;

perform one or more forward simulations based at least in part on the one or more candidate policies for the first vehicle and the distribution of probabilities at the current time for each second vehicle, wherein selecting the policy for the first vehicle is further based in part on results of the one or more forward simulations.

4. The system of claim 3, further wherein performing each of the one or more forward simulations includes:

selecting a sample policy for the first vehicle from the candidate policies;

selecting a sample policy for each of the one or more second vehicles from the respective sets of potential policies for each of the respective second vehicles; and

performing the forward simulation based on the selected sample policies for the first and one or more second vehicles.

5. The system of claim 4, wherein the computer is further programmed to:

identify one or more rewards reflecting a desired result;

wherein selecting the policy for the first vehicle is further based in part on the identified rewards.

6. The system of claim 5, wherein the computer is further programmed to:

compare a result of the one or more forward simulations with respect to each of the identified one or more rewards;

assign, based at least in part on the comparison, a weighted value to the result of the forward simulation with respect to each of the identified one or more rewards;

compute a sum of the weighted values of the result for each of the one or more forward simulations with respect to each of the identified one or more rewards; and

identify a forward simulation from the one or more forward simulations with a highest sum, wherein selecting the policy for the first vehicle is based at least in part on the selected sample policy used for the first vehicle in the identified forward simulation.

7. The system of claim 1, wherein the computer is further programmed to:

determine, based on the data, that the behavior of one of the one or more second vehicles cannot be explained by any policy from the set of pre-defined policies.

8. The system of claim 1, wherein identifying the change points includes inferring a maximum a posteriori set of change points at which change points between policies have occurred, based on the set of three or more pre-defined policies.

9. The system of claim 1, wherein the computer is further programmed to:

identify, from the first vehicle, the one or more second vehicles within a specified distance of the first vehicle, wherein the specified distance is based at least in part on a current traffic environment of the first vehicle.

10. The system of claim 9, wherein selecting the set of potential policies for each second vehicle is based at least in part on the current traffic environment.

11. A method comprising:

receiving data about a current and historical trajectory of each of one or more second vehicles;

selecting, from a set of three or more pre-defined policies, a set of potential policies for each second vehicle, each of the potential policies in each respective set of potential policies including a respective possible second vehicle trajectory and one or more rules for the respective second vehicle trajectory;

identifying a set of change points for each second vehicle'"'"'s historical trajectory, wherein each change point specifies a location and a time along the respective second vehicle historical trajectory at which it is determined that the respective second vehicle changed from a first policy from the set of potential policies for the respective second vehicle to a second policy from the set of potential policies for the respective second vehicle;

determining, for each segment between a respective, adjacent pair of change points along the respective historical trajectory for each second vehicle, a respective policy from the set of potential policies for the respective second vehicle that fits to the historical trajectory of the respective second vehicle;

identifying, based at least in part on the policy from the set of potential policies for each segment between the respective adjacent pair of change points that fits to the historical trajectory of each respective second vehicle, a distribution of probabilities for each second vehicle at a current time;

each probability of each respective distribution of probabilities associated with one of the potential policies from the set of potential policies for the respective second vehicle;

selecting a policy for the first vehicle from the set of three or more pre-defined policies, based at least in part on the respective distributions of probabilities at the current time for the respective one or more second vehicles; and

controlling the vehicle via an instruction provided by the computer to a controller based on the selected policy for the first vehicle.

12. The method of claim 11, wherein the selected policy for the first vehicle further indicates one or more alternative trajectories, each of the alternative trajectories associated with potential policies of the one or more second vehicles.

13. The method of claim 11, further comprising:

identifying one or more candidate policies for the first vehicle;

performing one or more forward simulations based at least in part on the one or more candidate policies for the first vehicle and the distribution of probabilities at the current time for each of the one or more second vehicles, wherein selecting the policy for the first vehicle is further based in part on results of the one or more forward simulations.

14. The method of claim 13, wherein performing each of the one or more forward simulations includes:

selecting a sample policy for the first vehicle from the candidate policies;

selecting a sample policy for each of the one or more second vehicles from the respective set of potential policies; and

performing the forward simulation based on the selected sample policies for the first and one or more second vehicles.

15. The method of claim 14, further comprising:

identifying one or more rewards reflecting a desired result;

wherein selecting the policy for the first vehicle is further based in part on the identified rewards.

16. The method of claim 15, further comprising:

comparing a result of the one or more forward simulations with respect to each of the identified one or more rewards;

assigning, based at least in part on the comparison, a weighted value to the result of the forward simulation with respect to each of the identified one or more rewards;

computing a sum of the weighted values of the result for each of the one or more forward simulations with respect to each of the identified one or more rewards; and

identifying a forward simulation from the one or more forward simulations with a highest sum, wherein selecting the policy for the first vehicle is based at least in part on the selected sample policy used for the first vehicle in the identified forward simulation.

17. The method of claim 11, further comprising:

determining, based on the data, that the behavior of one of the one or more second vehicles cannot be explained by any policy from the set of pre-defined policies.

18. The method of claim 11, wherein identifying the change points includes inferring a maximum a posteriori set of change points at which change points between policies have occurred, based on the set of three or more pre-defined policies.

19. The method of claim 11, further comprising:

identifying, from the first vehicle, the one or more second vehicles within a specified distance of the first vehicle, wherein the specified distance is based at least in part on a current traffic environment of the first vehicle.

20. The method of claim 19, wherein selecting the set of potential policies for each second vehicle is based at least in part on the current traffic environment.