Adaptive vehicle control system with driving style recognition based on vehicle passing maneuvers
An adaptive vehicle control system that classifies a driver's driving style based on vehicle passing maneuvers. A process determines whether the vehicle has started a lane change from a first lane to a second lane as a first phase of the passing maneuver. If so, the process determines whether the lane change to the second lane has been completed as a second phase of the passing maneuver. If the lane change to the second lane has been completed, the process determines whether there has been a lane change back to the first lane to complete the passing maneuver as a third phase of the passing maneuver. Further, the process determines whether a time threshold has been exceeded after the passing maneuver is in a second phase, but no lane change back to the first lane has occurred.
1. A method for identifying whether a vehicle maneuver is a passing maneuver, said method comprising:
reading a vehicle speed signal and a vehicle yaw rate signal from vehicle sensors;
determining whether the vehicle has started a lane change from a first lane to a second lane as a first phase of the passing maneuver;
determining whether the lane change to the second lane has been completed as a second phase of the passing maneuver; and
determining whether there has been a lane change back to the first lane to complete the passing maneuver if the passing maneuver is in the second phase as a third phase of the passing maneuver.
2. The method according to claim 1 further comprising determining whether a lane change has been aborted when the passing maneuver is in the first phase and the third phase.
3. The method according to claim 1 further comprising determining whether a time threshold has been exceeded after the passing maneuver is in the second phase, but no lane change back to the first lane has occurred.
4. The method according to claim 1 further comprising classifying the passing maneuver to determine a driver's driving style after the passing maneuver has been completed.
5. The method according to claim 4 wherein classifying the passing maneuver includes using selected discriminant features derived or obtained from the passing maneuver.
6. The method according to claim 5 wherein the discriminant features are derived or obtained from the group comprising a maximum vehicle yaw rate, a maximum vehicle lateral acceleration, a maximum lateral jerk, a distance for lane change, an average vehicle speed, a maximum speed variation, a maximum braking pedal force, a maximum throttle percentage, a minimum distance to a preceding vehicle, a maximum range rate to the preceding vehicle and a minimum distance to a following vehicle for the first phase of the passing maneuver.
7. The method according to claim 6 wherein the discriminant features are derive or obtained from the group comprising a maximum throttle percentage, an average throttle percentage, a distance traveled and maximum speed variation for the second phase of the passing maneuver.
8. The method according to claim 5 wherein a discriminant feature extraction method is selected from the group comprising principal component analysis, linear discriminant analysis, colonial principal component analysis and generalized discriminant analysis, and a discriminant feature selection method is obtained from the group comprising exhaustive search, branch-and-bound search, sequential forward/backward selection and sequential forward/backward floating search.
9. The method according to claim 4 wherein classifying the passing maneuver includes using support vector machines.
10. The method according to claim 4 wherein classifying the passing maneuver includes using a technique selected from a group comprising fuzzy logic, clustering, neural networks, support vector machines and threshold-based logic.
11. A method for identifying a vehicle maneuver as a passing maneuver, said method comprising:
reading a vehicle speed signal and a vehicle yaw rate signal form vehicle sensors;
determining whether the vehicle has started a lane change from a first lane to a second lane as a first phase of the passing maneuver;
determining whether the lane change to the second lane has been completed as a second phase of the passing maneuver;
determining whether there has been a lane change back to the first lane to complete the passing maneuver if the passing maneuver is in the second phase as a third phase of the passing maneuver;
determining whether a time threshold has been exceeded after the passing maneuver is in the second phase, but no lane change back to the first lane has occurred; and
classifying the passing maneuver to determine a driver's driving style after the passing maneuver has been completed.
12. The method according to claim 11 further comprising determining whether a lane change has been aborted when the passing maneuver is in the first phase and the third phase.
13. The method according to claim 11 wherein classifying the passing maneuver includes using selected discriminant features derived or obtained from the passing maneuver.
14. The method according to claim 13 wherein the discriminant features are derived or obtained from the group comprising a maximum vehicle yaw rate, a maximum vehicle lateral acceleration, a maximum lateral jerk, a distance for lane change, an average vehicle speed, a maximum speed variation, a maximum braking pedal force, a maximum throttle percentage, a minimum distance to a preceding vehicle, a maximum range rate to the preceding vehicle and a minimum distance to a following vehicle for the first phase of the passing maneuver.
15. The method according to claim 14 wherein the discriminant features are derived or obtained from the group comprising a maximum throttle percentage, an average throttle percentage, a distance traveled and maximum speed variation for the second phase of the passing maneuver.
16. The method according to claim 13 wherein a discriminant feature extraction method is selected from the group comprising principal component analysis, linear discriminant analysis, colonial principal component analysis and generalized discriminant analysis, and a discriminant feature selection method is obtained from the group comprising exhaustive search, branch-and-bound search, sequential forward/backward selection and sequential forward/backward floating search.
17. The method according to claim 11 wherein classifying the passing maneuver includes using support vector machines.
18. A system for identifying a vehicle maneuver as a passing maneuver, said system comprising:
vehicle sensors that provide a vehicle speed signal and a vehicle yaw rate signal;
means for determining whether the vehicle has started a lane change from a first lane to a second lane as a first phase of the passing maneuver;
means for determining whether the lane change to the second lane has been completed as a second phase of the passing maneuver; and
means for determining whether there has been a lane change back to the first lane to complete the passing maneuver if the passing maneuver is in the second phase as a third phase of the passing maneuver.
19. The system according to claim 18 further comprising means for determining whether a lane change has been aborted when the passing maneuver is in the first phase and the third phase.
20. The system according to claim 18 further comprising means for determining whether a time threshold has been exceeded after the passing maneuver is in the second phase, but no lane change back to the first lane has occurred.
21. The system according to claim 18 further comprising means for classifying the passing maneuver to determine a driver's driving style after the passing maneuver has been completed.
22. The system according to claim 21 wherein the means for classifying the passing maneuver includes means for extracting and selecting discriminant features from the passing maneuver.
23. The system according to claim 22 wherein the discriminant features are derive or obtained from the group comprising a maximum vehicle yaw rate, a maximum vehicle lateral acceleration, a maximum lateral jerk, a distance for lane change, an average vehicle speed, a maximum speed variation, a maximum braking pedal force, a maximum throttle percentage, a minimum distance to a preceding vehicle, a maximum range rate to the preceding vehicle and a minimum distance to a following vehicle for the first phase of the passing maneuver.
24. The system according to claim 23 wherein the discriminant features are selected from the group comprising a maximum throttle percentage, an average throttle percentage, a distance traveled and maximum speed variation for the second phase of the passing maneuver.
25. The system according to claim 22 wherein a discriminant feature extraction method is selected from the group comprising principal component analysis, linear discriminant analysis, colonial principal component analysis and generalized discriminant analysis, and a discriminant feature selection method is obtained from the group comprising exhaustive search, branch-and-bound search, sequential forward/backward selection and sequential forward/backward floating search.
26. The system according to claim 21 wherein the means for classifying the passing maneuver uses support vector machines.
27. A method for identifying a vehicle passing maneuver, said method comprising:
providing a vehicle speed signal from a speed sensor;
providing a vehicle yaw rate signal from a vehicle yaw rate sensor;
providing a steering angle signal from a steering angle sensor;
providing a heading angle signal from a heading angle sensor;
providing a vehicle acceleration signal from an acceleration sensor;
providing a vehicle position signal from a position sensor; and
using the sensor signals to determine whether the vehicle is in the passing maneuver.