Systems and methods for controlling longitudinal acceleration based on lateral objects

1. A method for controlling acceleration of a vehicle, comprising:

detecting, by a computing device, a lateral distance from a point on a trajectory of the vehicle to a first object the vehicle is expected to pass when following the trajectory, the first object being located off of and to a side of the trajectory and the point representing a future location of the vehicle while passing the first object;

selecting, by the computing device, whether acceleration limiting is to be performed by the vehicle based on the lateral distance;

obtaining an amount by which the acceleration of the vehicle is to be limited, when a selection is made that acceleration limiting is to be performed by the vehicle; and

causing, by the computing device, the vehicle to perform operations for autonomous driving with limiting of acceleration by the obtained amount.

2. The method according to claim 1 , wherein the obtaining comprises:

defining an interval with a first end located a pre-defined distance from the first object and a second opposing end located the pre-defined distance from the first object;

obtaining a distance between the first object and each index point of the trajectory within the interval;

obtaining a desired acceleration for each said index point of the trajectory based on the respective one of the distances; and

selecting the smallest value from the desired accelerations for limiting acceleration of the vehicle.

3. The method according to claim 2 , wherein the desired acceleration is obtained using a look-up table comprising a plurality of different acceleration values that are indexed by a plurality of different distance values.

4. The method according to claim 2 , further comprising adjusting the smallest value based on a class of the object.

5. The method according to claim 4 , wherein the smallest value is adjusted by a first amount when the object is associated with a first class and by a second different amount when the object is associated with a second different class.

6. The method according to claim 1 , wherein the vehicle is expected to additionally pass a second object and the obtaining further comprises:

for each of the first and second objects,

defining an interval with a first end located a pre-defined distance from the first or second object and a second opposing end located the pre-defined distance from the first or second object;

obtaining a distance between the first or second object and each index point of the trajectory within the interval;

obtaining a desired acceleration for each said index point of the trajectory based on the respective one of the distances;

selecting a smallest value from the desired accelerations for the first and second object at each index point of the trajectory; and

generating at least one resultant acceleration value using the selected smallest values.

7. The method according to claim 6 , wherein the at least one resultant acceleration value is generated by a machine learning algorithm that is trained to generate resultant acceleration values based on context information derived from sensor data.

8. The method according to claim 7 , wherein the context information comprises at least one of a relative location of vehicle and the first object, a relative location of the vehicle and the second object, a position of vehicle, a position of first object, a position of the second object, the trajectory, a predicted trajectory for the first object, a predicted trajectory for the second object, a classification of the first object, a classification of the second object, a direction of travel for the vehicle, a direction of travel for the first object and a direction of travel for the second object.

9. The method according to claim 6 , further comprising adjusting the at least one resultant acceleration value based on at least one of a classification of the first object and a classification of the second object.

10. The method according to claim 1 , wherein the selecting is further based on whether the first object is located entirely inside a non-drivable area for the vehicle and at an edge of a drivable area for the vehicle, whether the first object is located partially inside a drivable area for the vehicle, or whether the first object is a moving object biased in a drivable area for the vehicle.

11. A system, comprising:

a processor;

a non-transitory computer-readable storage medium comprising programming instructions that are configured to cause the processor to implement a method for controlling acceleration of a vehicle, wherein the programming instructions comprise instructions to:

detect a lateral distance from a point on a trajectory of the vehicle to a first object the vehicle is expected to pass when following the trajectory, the first object being located off of and to a side of the trajectory and the point representing a future location of the vehicle while passing the first object;

select whether acceleration limiting is to be performed by the vehicle based on the lateral distance;

obtain an amount by which the acceleration of the vehicle is to be limited, when a selection is made that acceleration limiting is to be performed by the vehicle; and

cause the vehicle to perform operations for autonomous driving with limiting of acceleration by the obtained amount.

12. The system according to claim 11 , wherein the amount by which the acceleration of the vehicle is to be limited is obtained by:

defining an interval with a first end located a pre-defined distance from the first object and a second opposing end located the pre-defined distance from the first object;

obtaining a distance between the first object and each index point of the trajectory within the interval;

obtaining a desired acceleration for each said index point of the trajectory based on the respective one of the distances; and

selecting the smallest value from the desired accelerations for limiting acceleration of the vehicle.

13. The system according to claim 12 , wherein the desired acceleration is obtained using a look-up table comprising a plurality of different acceleration values that are indexed by a plurality of different distance values.

14. The system according to claim 12 , wherein the programming instructions further comprise instructions to adjust the smallest value based on a class of the object.

15. The system according to claim 14 , wherein the smallest value is adjusted by a first amount when the object is associated with a first class and by a second different amount when the object is associated with a second different class.

16. The system according to claim 11 , wherein the vehicle is expected to additionally pass a second object and the amount by which the acceleration of the vehicle is to be limited is obtained by:

for each of the first and second objects,

defining an interval with a first end located a pre-defined distance from the first or second object and a second opposing end located the pre-defined distance from the first or second object;

obtaining a distance between the first or second object and each index point of the trajectory within the interval;

obtaining a desired acceleration for each said index point of the trajectory based on the respective one of the distances;

selecting a smallest value from the desired accelerations for the first and second object at each index point of the trajectory; and

generating at least one resultant acceleration value using the selected smallest values.

17. The system according to claim 16 , wherein the at least one resultant acceleration value is generated by a machine learning algorithm that is trained to generate resultant acceleration values based on context information derived from sensor data.

18. The system according to claim 17 , wherein the context information comprises at least one of a relative location of vehicle and the first object, a relative location of the vehicle and the second object, a position of vehicle, a position of first object, a position of the second object, the trajectory, a predicted trajectory for the first object, a predicted trajectory for the second object, a classification of the first object, a classification of the second object, a direction of travel for the vehicle, a direction of travel for the first object and a direction of travel for the second object.

19. The system according to claim 16 , wherein the programming instructions further comprise instructions to adjust the at least one resultant acceleration value based on at least one of a classification of the first object and a classification of the second object.

20. A non-transitory computer-readable medium that stores instructions that is configured to, when executed by at least one computing device, cause the at least one computing device to perform operations comprising:

detecting a lateral distance from a point on a trajectory of the vehicle to a first object the vehicle is expected to pass when following the trajectory, the first object being located off of and to a side of the trajectory and the point representing a future location of the vehicle while passing the first object;

selecting whether acceleration limiting is to be performed by the vehicle based on the lateral distance;

obtaining an amount by which the acceleration of the vehicle is to be limited, when a selection is made that acceleration limiting is to be performed by the vehicle; and

causing the vehicle to perform operations for autonomous driving with limiting of acceleration by the obtained amount.