Motion planning and intention prediction for autonomous driving in highway scenarios via graphical model-based factorization
Aspects of the disclosure are related to a method, apparatus, and system for planning a motion for a first vehicle, comprising: estimating past states of an observed second vehicle based on sensor inputs; predicting a future trajectory of the second vehicle based on the estimated past states; planning a future trajectory of the first vehicle based on the predicted future trajectory of the second vehicle and a safety cost function; and driving the first vehicle to follow the planned trajectory.
1. A method for planning a motion for a first vehicle, comprising:
by one or more processors, estimating past states of an observed second vehicle based on sensor inputs;
by the one or more processors, predicting a future trajectory of the second vehicle based on the estimated past states;
by the one or more processors, planning a future trajectory of the first vehicle based on the predicted future trajectory of the second vehicle and a safety cost function; and
by the one or more processors, driving the first vehicle to follow the planned trajectory,
wherein the planned trajectory includes an estimated lane change velocity.
2. The method of claim 1 , wherein the future trajectory of the second vehicle is modeled with a cubic spline comprising a 3rd order polynomial.
3. The method of claim 1 , wherein the predicting a future trajectory of the second vehicle is based further on a road structure.
4. The method of claim 1 , wherein the future trajectory of the first vehicle is modeled with a cubic spline comprising a 3rd order polynomial.
5. The method of claim 1 , wherein the future trajectory of the first vehicle is planned based further on at least one of a control smoothness factor, or a road structure, or any combination thereof.
6. The method of claim 1 , wherein the driving the first vehicle includes:
before a lane change, changing a velocity of the first vehicle from a start velocity to the estimated lane change velocity; and
after the lane change, changing the velocity of the first vehicle from the estimated lane change velocity to a target velocity.
7. A first vehicle apparatus for planning a motion, comprising:
one or more sensors configured to detect road and vehicle parameters;
a memory configured to store code and data;
a processor coupled to the memory, the processor configured to:
estimate past states of an observed second vehicle based on sensor inputs;
predict a future trajectory of the second vehicle based on the estimated past states;
plan a future trajectory of the first vehicle based on the predicted future trajectory of the second vehicle and a safety cost function; and
drive the first vehicle to follow the planned trajectory,
wherein the planned trajectory includes an estimated lane change velocity.
8. The first vehicle apparatus of claim 7 , wherein the future trajectory of the second vehicle is modeled with a cubic spline comprising a 3rd order polynomial.
9. The first vehicle apparatus of claim 7 , wherein the predicting a future trajectory of the second vehicle is based further on a road structure.
10. The first vehicle apparatus of claim 7 , wherein the future trajectory of the first vehicle is modeled with a cubic spline comprising a 3rd order polynomial.
11. The first vehicle apparatus of claim 7 , wherein the future trajectory of the first vehicle is planned based further on at least one of a control smoothness factor, or a road structure, or any combination thereof.
12. The first vehicle apparatus of claim 7 , wherein the driving the first vehicle includes:
before a lane change, changing a velocity of the first vehicle from a start velocity to the estimated lane change velocity; and
after the lane change, changing the velocity of the first vehicle from the estimated lane change velocity to a target velocity.
13. An apparatus for planning a motion for a first vehicle, comprising:
means for estimating past states of an observed second vehicle based on sensor inputs;
means for predicting a future trajectory of the second vehicle based on the estimated past states;
means for planning a future trajectory of the first vehicle based on the predicted future trajectory of the second vehicle and a safety cost function; and
means for driving the first vehicle to follow the planned trajectory,
wherein the planned trajectory includes an estimated lane change velocity.
14. The apparatus of claim 13 , wherein the future trajectory of the second vehicle is modeled with a cubic spline comprising a 3rd order polynomial.
15. The apparatus of claim 13 , wherein the predicting a future trajectory of the second vehicle is based further on a road structure.
16. The apparatus of claim 13 , wherein the future trajectory of the first vehicle is modeled with a cubic spline comprising a 3rd order polynomial.
17. The apparatus of claim 13 , wherein the future trajectory of the first vehicle is planned based further on at least one of a control smoothness factor, or a road structure, or any combination thereof.
18. The apparatus of claim 13 , wherein the means for driving the first vehicle includes:
means for changing a velocity of the first vehicle, before a lane change, from a start velocity to the estimated lane change velocity; and
means for changing the velocity of the first vehicle, after the lane change, from the estimated lane change velocity to a target velocity.
19. A non-transitory computer-readable medium comprising code which, when executed by a processor, causes the processor to perform a method comprising:
estimating past states of an observed second vehicle based on sensor inputs;
predicting a future trajectory of the second vehicle based on the estimated past states;
planning a future trajectory of the first vehicle based on the predicted future trajectory of the second vehicle and a safety cost function; and
driving the first vehicle to follow the planned trajectory,
wherein the planned trajectory includes an estimated lane change velocity.
20. The non-transitory computer-readable medium of claim 19 , wherein the future trajectory of the second vehicle is modeled with a cubic spline comprising a 3rd order polynomial.
21. The non-transitory computer-readable medium of claim 19 , wherein the predicting a future trajectory of the second vehicle is based further on a road structure.
22. The non-transitory computer-readable medium of claim 19 , wherein the future trajectory of the first vehicle is modeled with a cubic spline comprising a 3rd order polynomial.
23. The non-transitory computer-readable medium of claim 19 , wherein the future trajectory of the first vehicle is planned based further on at least one of a control smoothness factor, or a road structure, or any combination thereof.
24. The non-transitory computer-readable medium of claim 19 , wherein the driving the first vehicle includes:
before a lane change, changing a velocity of the first vehicle from a start velocity to the estimated lane change velocity; and
after the lane change, changing the velocity of the first vehicle from the estimated lane change velocity to a target velocity.