Driving action estimating device, driving support device, vehicle evaluating system, driver model creating device, and driving action determining device

• US 8,140,241 B2
• Filed: 12/27/2006
• Issued: 03/20/2012
• Est. Priority Date: 12/28/2005
• Status: Active Grant

• Current Assignee: Equos Research Co., Ltd. (Aisin Corp.), Nagoya University
• Original Assignee: Equos Research Co., Ltd. (Aisin Corp.), Nagoya University
• Inventors: Takeda, Kazuya, Itou, Katunobu, Nomoto, Hirokazu, Fujii, Kazuaki, Suzuki, Seiichi, Miyajima, Chiyomi, Ozawa, Koji
• Primary Examiner(s): Zanelli, Michael J.
• Application Number:
  US12/087,130
• Publication Number:
  US 20090234552A1
• Time in Patent Office: 1,910 Days
• Field of Search: None
• US Class Current: 701/93
• CPC Class Codes: B60W 30/16 Control of distance between vehicles, e.g. keeping a distance to preceding vehicle
• CPC Class Codes: B60W 40/09 Driving style or behaviour
• CPC Class Codes: G08G 1/166 for active traffic, e.g. moving vehicles, pedestrians, bikes
• CPC Class Codes: G08G 1/167 Driving aids for lane monitoring, lane changing, e.g. blind spot detection

Abstract

A driver model with higher precision is created as an evaluation standard for a driving condition in a normal condition. By detecting biometric information of a driver, whether a driver is in a usual condition or not is recognized. Then, data of driving conditions are collected while the driver is driving, and from the driving condition data, a part indicating that the driver operates in a usual condition is extracted to create a driver model. Further, the driver model is created taking only a case of driving in a normal condition as a driving action in normal times based on biometric information of the driver, and hence the driver model becomes more precise and neutral. Further, by using a GMM (Gaussian mixture model) for the driver model, a driver model for each driver can be created easily, and moreover, by calculation to maximize a conditional probability, a driving operation action is easily estimated and outputted.

Claims

1. A driving action estimating device, comprising:

a driver model for accelerator pedal operation and a driver model for brake pedal operation, each driver model being in the form of a Gaussian mixture model calculated by an EM algorithm, with time-series data of N types of characteristic amounts detected during travel of a vehicle as learning data, as a probability distribution of existence of the respective learning data in N dimensional spaces;

characteristic amount obtaining means for obtaining at least one or more characteristic amounts from the N types excluding a certain characteristic amount x;

maximum posterior probability calculating means for calculating a maximum posterior probability in the driver model for the obtained characteristic amount; and

output means for outputting an estimated value of the certain characteristic amount x for the obtained characteristic amount based on the calculated maximum posterior probability.

2. The driving action estimating device according to claim 1, wherein the N types of characteristic amounts include a time change amount for n types (n<

N) of characteristic amounts.

3. The driving action estimating device according to claim 1, wherein the characteristic amount x includes an operation amount of an operation device operated directly by a driver and a time change amount of the operation amount.

4. A driving support device, comprising:

a driving action estimating device according to claim 1, wherein the learning data comprises an accelerator operation amount, a brake operation amount, a vehicle speed of the driver'"'"'s vehicle, and an inter-vehicle distance between the driver'"'"'s vehicle and a leading vehicle immediately ahead of the driver'"'"'s vehicle as characteristic amounts, and estimates an accelerator operation amount and a brake operation amount as the characteristic amounts x;

traveling data obtaining means for obtaining the vehicle speed of the driver'"'"'s vehicle and the inter-vehicle distance;

traveling control means for performing automatic cruise control with respect to the leading vehicle by controlling an engine throttle and a brake pedal according to an accelerator operation amount and a brake operation amount that are estimated with respect to the obtained traveling data by the driving action estimating device.

5. A vehicle evaluating system, comprising:

a driving action estimating device according to claim 1, wherein the learning data comprises an accelerator operation amount, a brake operation amount, a vehicle speed of the driver'"'"'s vehicle, and an inter-vehicle distance between the driver'"'"'s vehicle and a leading vehicle immediately ahead of the driver'"'"'s vehicle as characteristic amounts, and estimates an accelerator operation amount and a brake operation amount as the characteristic amounts x;

means for obtaining vehicle performance data for the driver'"'"'s vehicle as a target of evaluation;

means for obtaining traveling data and a traveling route model for simulation;

vehicle dynamics calculating means for estimating a behavior of the driver'"'"'s vehicle including acceleration of the driver'"'"'s vehicle as a target of evaluation with respect to an accelerator operation amount and a brake operation amount that are obtained by inputting the obtained traveling data and traveling route model to the driving action estimating device; and

evaluating means for evaluating traveling performance of the driver'"'"'s vehicle as a target of evaluation from the estimated behavior of the vehicle.

6. A driving support device, comprising:

driver model obtaining means for obtaining a driver model for accelerator pedal operation and a driver model for brake pedal operation under normal conditions, each driver model being in the form of a Gaussian mixture model calculated by an EM algorithm, with time-series data of N types of characteristic amounts detected during travel of a vehicle as learning data, as a probability distribution of existence of the respective learning data in N dimensional spaces;

driving operation estimating means for estimating a driving operation which is usually operated in a normal condition using the obtained driver model;

driving action determining means for determining a driving action of a driver from the estimated driving operation and a driving operation based on current driving operation information; and

driving support means for providing driving support according to the determined driving action.

7. The driving support device according to claim 6, wherein the driver model obtaining means obtains a driver model corresponding to a current traveling environment from driver models of driving operations in a normal condition created for each traveling environment.

8. The driving support device according to claim 6 further comprising:

driver condition determining means for determining a condition of a driver from biometric information of the driver,

wherein the driving support means performs a driving support according to the determined driving action and the determined driver condition.

9. The driving support device according to claim 6, wherein the driving support means performs at least one driving support function selected from the group consisting of warnings by voice or image, providing information, vibration, and guidance to a rest station.

10. A driving action determining device, comprising:

driver model obtaining means for obtaining a driver model for accelerator pedal operation and a driver model for brake pedal operation under normal conditions, each driver model being in the form of a Gaussian mixture model calculated by an EM algorithm, with time-series data of N types of characteristic amounts detected during travel of a vehicle as learning data, as a probability distribution of existence of the respective learning data in N dimensional spaces;

driving operation estimating means for estimating, as driving operations, a brake pedal operation and an accelerator pedal operation, which operations are usually made under the normal conditions, using the obtained driver models; and

driving action determining means for determining a brake pedal operation and an accelerator operation of a driver from the estimated driving operations and current driving operations based on current driving operation information.