System and method for learning driver preference and adapting lane centering controls to driver behavior

1. A method of operating a vehicle, comprising:

navigating the vehicle over a road segment using a driving behavior of the vehicle;

receiving, at a processor, a driver's behavior of a driver of the vehicle for the road segment as the vehicle navigates the road segment;

constructing a knowledge matrix by learning the driver's behavior for a plurality of navigations of the road segment;

creating, at the processor, a reward function based on a difference between the driver's behavior and the driving behavior of the vehicle;

updating, at the processor, the knowledge matrix based on the reward function;

selecting, at the processor, the behavior policy from the updated knowledge matrix; and

controlling, via a processor, the vehicle to navigate the road segment using the behavior policy.

2. The method of claim 1 , wherein the knowledge matrix is based on at least one of the environmental state, a vehicle state, and a driver state.

3. The method of claim 1 , wherein learning the driver's behavior further comprises measuring at least one of a vehicle speed selected by the driver, a lateral control of the vehicle selected by the driver, and an acceleration or deceleration of the vehicle selected by driver when the driver navigates the vehicle over the road segment.

4. The method of claim 1 , wherein the threshold associated with the road segment includes a safety limit of the road segment and creating the behavior policy includes modifying the learned driver's behavior based on the safety limit of the road segment.

5. The method of claim 1 , wherein the driver's behavior further comprises at least one of: the driver's behavior within a lane of the road segment; and the driver's behavior for changing lanes in the road segment.

6. The method of claim 1 , further comprising learning the driver's behavior in one of an offline learning mode in which the vehicle is driven by the driver and an online learning mode in which the vehicle is controlled by the processor as the driver operates a control of the vehicle.

7. A system for operating a vehicle, comprising:

a processor configured to:

navigate the vehicle over a road segment using a driving behavior of the vehicle;

receive input of a driver's behavior for navigating the vehicle over a road segment as the vehicle navigates the road segment;

construct a knowledge matrix by learning the driver's behavior for a plurality of navigations of the road segment;

create a reward function based on a difference between the driver's behavior and the driving behavior of the vehicle;

update the knowledge matrix based on the reward function;

select the behavior policy from the updated knowledge matrix; and

control the vehicle to navigate the road segment using the behavior policy.

8. The system of claim 7 , wherein the knowledge matrix is based on at least one of the environmental state, a vehicle state, and a driver state.

9. The system of claim 7 , wherein the processor is further configured to learn the driver's behavior by measuring at least one of a vehicle speed selected by the driver, a lateral control of the vehicle selected by the driver, and an acceleration or deceleration of the vehicle selected by the driver when the driver navigates the vehicle over the road segment.

10. The system of claim 7 , wherein the threshold associated with the road segment includes a safety limit of the road segment, and the processor is configured to create the behavior policy by modifying the learned driver's behavior based on the safety limit of the road segment.

11. The system of claim 7 , further comprising an electronics package transferable to and from the vehicle.

12. The system of claim 7 , wherein the processor is further configured to learn the driver's behavior in one of an offline learning mode in which the vehicle is driven by the driver and an online learning mode in which the vehicle is controlled by the processor as the driver operates a control of the vehicle.

13. A vehicle, comprising:

a processor configured to:

navigate the vehicle over a road segment using a driving behavior of the vehicle;

receive input of a driver's behavior of a driver of the vehicle as for the road segment as the vehicle navigates the road segment;

construct a knowledge matrix by learning the driver's behavior for a plurality of navigations of the road segment;

create a reward function based on a difference between the driver's behavior and the driving behavior of the vehicle;

update the knowledge matrix based on the reward function;

select the behavior policy from the updated knowledge matrix; and

control the vehicle to navigate the road segment using the behavior policy.

14. The vehicle of claim 13 , further comprising an additional processor that is transferable to and from the vehicle, the additional processor configured to perform at least one of: (i) modifying the path planning commands based on the learned driver's behavior; and (ii) adapting a lane centering control command to the learned-driver-behavior.

15. The vehicle of claim 13 , wherein the processor is further configured to learn the driver's behavior by measuring at least one of a vehicle speed selected by the driver, a lateral control of the vehicle selected by the driver, and an acceleration or deceleration of the vehicle selected by the driver when the driver navigates the vehicle over the road segment.

16. The vehicle of claim 13 , wherein the threshold associated with the road segment includes a safety limit of the road segment, and the processor is configured to create the behavior policy by modifying the learned driver's behavior based on the safety limit of the road segment.

17. The vehicle of claim 13 , wherein the processor is further configured to learn the driver's behavior in one of an offline learning mode in which the vehicle is driven by the driver and an online learning mode in which the vehicle is controlled by the processor as the driver operates a control of the vehicle.