Systems and methods for trajectory determination

• US 10,809,731 B2
• Filed: 12/28/2018
• Issued: 10/20/2020
• Est. Priority Date: 07/13/2017
• Status: Active Grant

• Current Assignee: Beijing Voyager Technology Co., Ltd.
• Original Assignee: Beijing Voyager Technology Co., Ltd.
• Inventors: Luo, Wei
• Primary Examiner(s): Ahmed, Masud
• Application Number:
  US16/236,313
• Publication Number:
  US 20190155290A1
• Time in Patent Office: 662 Days
• CPC Class Codes: B60W 30/00 Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units
• CPC Class Codes: G01C 21/3407 specially adapted for specific applications
• CPC Class Codes: G05D 1/0088 characterized by the autonomous decision making process, e.g. artificial intelligence, predefined behaviours using knowledge based models G06N5/00
• CPC Class Codes: G05D 1/0214 in accordance with safety or protection criteria, e.g. avoiding hazardous areas monitoring the location of vehicles within a certain area, e.g. forbidden or allowed areas, in traffic control systems for road vehicles G08G1/13
• CPC Class Codes: G05D 1/0217 in accordance with energy consumption, time reduction or distance reduction criteria
• CPC Class Codes: G05D 1/0274 using mapping information stored in a memory device navigation using map-matching G01C21/30

Abstract

Systems and methods for trajectory determination are provided. The systems perform the operations including obtaining status information of a vehicle; determining one or more original trajectories from a current location of the vehicle to a destination based on the status information of the vehicle; determining one or more candidate trajectories from the original trajectories; and determining a target trajectory from the candidate trajectories based on a trajectory evaluation model.

Claims

1. A system for determining a trajectory for a vehicle, comprising:

at least one storage medium including a set of instructions; and

at least one processor configured to communicate with the at least one storage medium, wherein when executing the set of instructions, the at least one processor is directed to;

obtain status information of a vehicle;

obtain a current location of the vehicle and a destination;

determine one or more sub-target locations related to the destination, wherein a count of the one or more sub-target locations are determined based on the status information of the vehicle;

determine one or more original trajectories from the current location of the vehicle to the destination based on the status information of the vehicle and the one or more sub-target locations;

determine one or more candidate trajectories from the one or more original trajectories based on results of a collision detection test, wherein each of the one or more candidate trajectories has no obstacle, as determined by the at least one processor;

obtain an trajectory evaluation model;

extract feature vectors from the one or more candidate trajectories, each feature vector corresponding to a candidate trajectory and including a speed of the vehicle and a similarity for the candidate trajectory between a current moment T+1 and a previous moment T; and

determine a target trajectory from the one or more candidate trajectories based on the trajectory evaluation model and the feature vectors for the one or more candidate trajectories.

2. The system of claim 1, wherein the status information of the vehicle includes at least one of:

a driving direction of the vehicle,

the speed of the vehicle,

an accelerated speed of the vehicle,

a wheel base of the vehicle, or

a front wheel steering angle of the vehicle.

3. The system of claim 1, wherein to determine the one or more sub-target locations related to the destination, the at least one processor is further directed to:

determine the one or more sub-target locations related to the destination based on a current speed of the vehicle.

4. The system of claim 1, wherein to determine the one or more candidate trajectories from the one or more original trajectories, the at least one processor is further directed to:

determine one or more obstacles in a grid map;

perform a collision detection between the one or more original trajectories and the one or more obstacles; and

determine the one or more candidate trajectories from the one or more original trajectories based on the collision detection.

5. The system of claim 1, wherein the trajectory model is a machine learning model and the machine learning model is trained and updated in real time.

6. The system of claim 1;

wherein the at least one processor is further directed to;

obtain a plurality of sample trajectories;

determine training data based on the plurality of sample trajectories; and

determine the trajectory evaluation model based on the training data.

7. The system of claim 6;

wherein to determine the training data based on the plurality of sample trajectories, the at least one processor is further directed to;

extract one or more sample features with respect to the plurality of sample trajectories and a sample evaluation result associated with each of the plurality of sample trajectories; and

determine the one or more sample features and the corresponding sample evaluation results as the training data.

8. The system of claim 7, wherein the one or more sample features include at least one of:

a sample starting location,

a sample destination,

a sample speed of a vehicle,

a sample accelerated speed of a vehicle, or

a sample curvature.

9. The system of claim 1, wherein the at least one processor is further directed to;

determine a speed planning strategy based on the target trajectory.

10. The system of claim 1, wherein to determine the one or more sub-target locations related to the destination, the at least one processor is further directed to:

determine the one or more sub-target locations related to the destination by randomly sampling in a target area, wherein the target area covers a centerline of a lane.

15. The method of claim 1, wherein the determining the one or more sub-target locations related to the destination comprises:

determining the one or more sub-target locations related to the destination by randomly sampling in a target area, wherein the target area covers a centerline of a lane.

11. A method implemented on a computer device having a processor, a storage medium, and a communication platform connected to a network, the method comprising:

obtaining, by the processor, status information of a vehicle;

obtaining, by the processor, a current location of the vehicle and a destination;

determining, by the processor, one or more sub-target locations related to the destination, wherein a count of the one or more sub-target locations are determined based on the status information of the vehicle;

determining, by the processor, one or more original trajectories from the current location of the vehicle to the destination based on the status information of the vehicle and the one or more sub-target locations;

determining, by the processor, one or more candidate trajectories from the one or more original trajectories based on result of a collision detection test, wherein each of the one or more candidate trajectories has no obstacle, as determined by the at least one processor;

obtaining, by the processor, an trajectory evaluation model;

extracting, by the processor, feature vectors from the one or more candidate trajectories, each feature vector corresponding to a candidate trajectory and including a speed of the vehicle and a similarity for the candidate trajectory between a current moment T+1 and a pervious moment T; and

determining, by the processor, a target trajectory from the one or more candidate trajectories based on the trajectory evaluation model and the feature vectors for the one or a ore available trajectories.

12. The method of claim 11, wherein the status information of the vehicle includes at least one of:

a driving direction of the vehicle,

the speed of the vehicle,

an accelerated speed of the vehicle,

a wheel base of the vehicle, or

a front wheel steering angle of the vehicle.

13. The method of claim 11, wherein the determining the one or more sub-target locations related to the destination comprises:

determining the one or more sub-target locations related to the destination based on a current speed of the vehicle.

14. The method of claim 11, wherein the determining the one or more candidate trajectories comprises:

determining one or more obstacles in a grid map;

performing a collision detection between the one or more original trajectories and the one or more obstacles; and

determining the one or more candidate trajectories from the one or more original trajectories based on the collision detection.

16. The method of claim 11, further comprising:

obtaining a plurality of sample trajectories;

determining training data based on the plurality of sample trajectories; and

determining the trajectory evaluation model based on the training data.

17. The method of claim 16, wherein the determining training data based on the plurality of sample trajectories comprises:

extracting one or more sample features with respect to the plurality of sample trajectories and a sample evaluation result associated with each of the plurality of sample trajectories; and

determining the one or more sample features and the corresponding sample evaluation results as the training data.

18. The method of claim 11, wherein the one or more sample features include at least one of:

a sample starting location,

a sample destination,

a sample speed of a vehicle,

a sample accelerated speed of a vehicle, or

a sample curvature of a sample trajectory.

19. The method of claim 11, further comprising:

determining a speed planning strategy based on the target trajectory.

20. A non-transitory computer readable medium, comprising at least one set of instructions for determining a trajectory for a vehicle, wherein when executed by at least one processor of an electronic terminal, the at least one set of instructions directs the at least one processor to perform acts of:

obtaining status information of a vehicle;

obtaining a current location of the vehicle and a destination;

determining one or more sub-target locations related to the destination, wherein a count of the one or more sub-target locations are determined based on the status information of the vehicle;

determining one or more original trajectories from the current location of the vehicle to the destination based on the status information of the vehicle and the one or more sub-target locations;

determining one or more candidate trajectories from the one or more original trajectories, bases on a collision detection test, wherein each of the one or more candidate trajectories has no obstacle, as determined by the at least one processor;

extract feature vectors form the one or more candidate trajectories;

each feature vector corresponding to a candidate trajectory and including a speed of the vehicle and a similarity for the candidate trajectory between a current moment T+1 and a pervious moment T; and

determining a target trajectory from the one or more candidate trajectories based on a trajectory evaluation model and the feature vectors for the one or more candidate trajectories.