Behavior control device and behavior control method for autonomous vehicles

1. A behavior control device for an autonomous vehicle, the behavior control device comprising:

a learning device configured to perform deep learning of a behavior pattern of a vehicle according to a driving environment;

a controller configured to control a behavior of the autonomous vehicle based on a result of the learning of the learning device,

wherein the learning device is configured to perform the deep learning of the behavior pattern of the vehicle using combined data in which various pieces of sensor data are combined, a steering wheel angle of the autonomous vehicle, and path information on a precise map.

2. The behavior control device of claim 1 , wherein the learning device further comprises:

a first learning device configured to perform deep learning of a behavior pattern of a preceding vehicle using the combined data.

3. The behavior control device of claim 2 , wherein the first learning device is configured to:

perform the deep learning of the behavior pattern of the preceding vehicle using at least one of a speed value of the preceding vehicle, brake lamp blinking information of the preceding vehicle, a speed value of a surrounding vehicle, or information on presence or absence of an intervening vehicle that is included in the combined data.

4. The behavior control device of claim 2 , wherein the controller is configured to:

predict the behavior pattern of the preceding vehicle by applying the combined data according to a driving environment at a current point to a learning result of the first learning device; and

derive a speed control value of the autonomous vehicle from the predicted behavior pattern of the preceding vehicle at a reference point.

5. The behavior control device of claim 4 , wherein the controller is configured to:

derive the speed control value of the autonomous vehicle when the autonomous vehicle drives at a speed less than a first reference speed.

6. The behavior control device of claim 1 , wherein the learning device further comprises:

a second learning device configured to perform deep learning of the behavior pattern of the autonomous vehicle using the combined data, the steering wheel angle of the autonomous vehicle, and the path information on the precise map.

7. The behavior control device of claim 6 , wherein the second learning device is configured to:

perform deep learning of the behavior pattern of the autonomous vehicle using a lane offset value of the autonomous vehicle included in the combined data.

8. The behavior control device of claim 6 , wherein the controller is configured to:

predict the behavior pattern of the autonomous vehicle by applying the combined data according to a driving environment at a current point, a steering wheel angle of the autonomous vehicle at the current point, and path information on the precise map at the current point to a learning result of the second learning device; and

derive a steering control value of the autonomous vehicle from the predicted behavior pattern of the autonomous vehicle at a reference period.

9. The behavior control device of claim 8 , wherein the controller is configured to:

derive the steering control value of the autonomous vehicle when the autonomous vehicle drives at a speed greater than a second reference speed.

10. A behavior control method for an autonomous vehicle, the behavior control method comprising:

performing, by a learning device, deep learning of a behavior pattern of a vehicle according to a driving environment; and

controlling, by a controller, a behavior of the autonomous vehicle based on a result of the learning of the learning device,

wherein performing the deep learning of the behavior pattern of the vehicle comprises:

performing, by the learning device, the deep learning of the behavior pattern of the vehicle using combined data in which various sensor data are combined, a steering wheel angle of the autonomous vehicle, and path information on a precise map.

11. The behavior control method of claim 10 , wherein performing the deep learning of the behavior pattern of the vehicle further comprises:

performing, by a first learning device, deep learning of a behavior pattern of a preceding vehicle using the combined data; and

performing, by a second learning device, deep learning of a behavior pattern of the autonomous vehicle using the combined data, the steering wheel angle of the autonomous vehicle, and the path information on the precise map.

12. The behavior control method of claim 11 , wherein performing the deep learning of the behavior pattern of the preceding vehicle further comprises:

performing the deep learning of the behavior pattern of the preceding vehicle using at least one of a speed value of the preceding vehicle, brake lamp blinking information of the preceding vehicle, a speed value of a surrounding vehicle, or information on presence or absence of an intervening vehicle that is included in the combined data.

13. The behavior control method of claim 11 , wherein controlling the behavior of the autonomous vehicle further comprises:

predicting the behavior pattern of the preceding vehicle by applying the combined data according to a driving environment at a current point to a learning result of the first learning device; and

deriving a speed control value of the autonomous vehicle from the predicted behavior pattern of the preceding vehicle at a reference period.

14. The behavior control method of claim 13 , wherein controlling the behavior of the autonomous vehicle comprises:

when the autonomous vehicle drives at a speed less than a first reference speed, controlling the behavior of the autonomous vehicle.

15. The behavior control method of claim 11 , wherein performing the deep learning of the behavior pattern of the autonomous vehicle further comprises:

performing the deep learning of the behavior pattern of the autonomous vehicle using a lane offset value of the autonomous vehicle that is included in the combined data.

16. The behavior control method of claim 11 , wherein controlling the behavior of the autonomous vehicle comprises:

predicting the behavior pattern of the autonomous vehicle by applying the combined data according to a driving environment at a current point, a steering wheel angle of the autonomous vehicle at the current point, and path information on the precise map at the current point to a learning result of the second learning device; and

deriving a steering control value of the autonomous vehicle from the predicted behavior pattern of the autonomous vehicle at a reference period.

17. The behavior control method of claim 16 , wherein controlling the behavior of the autonomous vehicle comprises:

when the autonomous vehicle drives at a speed greater than a second reference speed, controlling the behavior of the autonomous vehicle.