System and method for path planning of autonomous vehicles based on gradient
A system and method for path planning of autonomous vehicles based on gradient are disclosed. A particular embodiment includes: generating and scoring a first suggested trajectory for an autonomous vehicle; generating a trajectory gradient based on the first suggested trajectory; generating and scoring a second suggested trajectory for the autonomous vehicle, the second suggested trajectory being based on the first suggested trajectory and a human driving model; and outputting the second suggested trajectory if the score corresponding to the second suggested trajectory is within a score differential threshold relative to the score corresponding to the first suggested trajectory.
1. A system comprising:
a data processor; and
a path planning module, executable by the data processor, the path planning module being configured to:
generate and score a first suggested trajectory for an autonomous vehicle, the first suggested trajectory being defined by a first plurality of waypoints;
generate a trajectory gradient based on the first suggested trajectory, the trajectory gradient defining scale and direction data corresponding to each of the first plurality of waypoints of the first suggested trajectory, the scale and direction data defining a second plurality of waypoints;
generate and score a second suggested trajectory defined by the second plurality of waypoints for the autonomous vehicle; and
cause the autonomous vehicle to follow a path conforming to the second suggested trajectory if the score corresponding to the second suggested trajectory is within a score differential threshold relative to the score corresponding to the first suggested trajectory.
2. The system of claim 1 wherein the path planning module comprises machine learnable components.
3. The system of claim 1 wherein the second suggested trajectory is generated based on a human driving model, wherein the path planning module is configured to modify parameters of the human driving model based on the trajectory gradient.
4. The system of claim 1 wherein the trajectory gradient is generated by taking a mathematical derivative of a function defining the first suggested trajectory.
5. The system of claim 4 wherein the scale and direction data corresponding to each of the first plurality of waypoints are derived by taking the mathematical derivative of the function defining the first suggested trajectory.
6. The system of claim 1 wherein the second suggested trajectory is output to an autonomous vehicle control subsystem of the autonomous vehicle causing the autonomous vehicle to follow the second suggested trajectory.
7. A method comprising:
generating and scoring a first suggested trajectory for an autonomous vehicle, the first suggested trajectory being defined by a first plurality of waypoints;
generating a trajectory gradient based on the first suggested trajectory, the trajectory gradient defining scale and direction data corresponding to each of the first plurality of waypoints of the first suggested trajectory, the scale and direction data defining a second plurality of waypoints;
generating and scoring a second suggested trajectory defined by the second plurality of waypoints for the autonomous vehicle; and
causing the autonomous vehicle to follow a path conforming to the second suggested trajectory if the score corresponding to the second suggested trajectory is within a score differential threshold relative to the score corresponding to the first suggested trajectory.
8. The method of claim 7 further comprising:
obtaining a human driving model trained with training data corresponding to human driving behaviors, the human driving model being a machine learning system.
9. The method of claim 8 wherein the trajectory gradient is generated based on the first suggested trajectory and the human driving model.
10. The method of claim 7 further comprising:
outputting the second suggested trajectory if the score corresponding to the second suggested trajectory is within the score differential threshold relative to the score corresponding to the first suggested trajectory.
11. The method of claim 10 further comprising:
providing an autonomous vehicle control subsystem in data communication with a data processor, a path planning module, and vehicle control devices, the autonomous vehicle control subsystem receiving an output of the second suggested trajectory from the path planning module and manipulating the vehicle control devices to cause the autonomous vehicle to follow the path conforming to the second suggested trajectory.
12. The method of claim 7 wherein the trajectory gradient is generated based on the first suggested trajectory and a human driving model trained with training data corresponding to human driving behaviors, wherein the human driving behaviors comprise changing a speed or a direction of a vehicle.
13. A non-transitory machine-useable storage medium embodying instructions which, when executed by a machine, cause the machine to:
generate and score a first suggested trajectory for an autonomous vehicle, the first suggested trajectory being defined by a first plurality of waypoints;
generate a trajectory gradient based on the first suggested trajectory, the trajectory gradient defining scale and direction data corresponding to each of the first plurality of waypoints of the first suggested trajectory, the scale and direction data defining a second plurality of waypoints;
generate and score a second suggested trajectory defined by the second plurality of waypoints for the autonomous vehicle; and
cause the autonomous vehicle to follow a path conforming to the second suggested trajectory if the score corresponding to the second suggested trajectory is within a score differential threshold relative to the score corresponding to the first suggested trajectory.
14. The non-transitory machine-useable storage medium of claim 13 wherein the second suggested trajectory is generated based on a human driving model, wherein the human driving model is configured to retain information corresponding to human driving behaviors as mathematical or data representations.
15. The non-transitory machine-useable storage medium of claim 14 wherein at least part of the information corresponding to the human driving behaviors retained in the human driving model can be adjusted or expunged.
16. The non-transitory machine-useable storage medium of claim 15 wherein the information corresponding to the human driving behaviors expunged comprises driving with abrupt acceleration rates.
17. The non-transitory machine-useable storage medium of claim 15 wherein the information corresponding to the human driving behaviors comprises a relationship between a braking percentage and a time, the relationship between the braking percentage and the time being learned and retained as a smooth data curve.
18. The non-transitory machine-useable storage medium of claim 14 wherein the information corresponding to the human driving behaviors is encoded as a neural network or rules sets.
19. The non-transitory machine-useable storage medium of claim 13 wherein the instructions further cause the machine to: combine the score of the first suggested trajectory with the score of the second suggested trajectory to generate a gradient using a gradient descent process.
20. The non-transitory machine-useable storage medium of claim 13 wherein the second suggested trajectory is generated based on the trajectory gradient and a human driving model, wherein if a score difference between the first suggested trajectory and the second suggested trajectory is equal or higher than the score differential threshold, parameters of the human driving model are modified.